université de montréal détection des gènes de résistance ... · existence of associations...
TRANSCRIPT
27t1. 3/I.l
Université de Montréal
Détection des gènes de résistance aux agents antimicrobiens chez des souches
d’Escherichia cou entérotoxinogènes et extraintestinales
par
CHifiSTINE MAYNARD
Département de pathologie et microbiologie
Faculté de médecine vétérinaire
Mémoire présenté à la Faculté des études supérieures
en vue de l’obtention du grade
de Maître ès sciences (M. Sc.)
en sciences vétérinaires
option microbiologie
Avril, 2003
© Christine Maynard, 2003
f I-’
L.
Université dl1de Montréal
Direction des bibliothèques
AVIS
L’auteur a autorisé l’Université de Montréal à reproduire et diffuser, en totalitéou en partie, par quelque moyen que ce soit et sur quelque support que cesoit, et exclusivement à des fins non lucratives d’enseignement et derecherche, des copies de ce mémoire ou de cette thèse.
L’auteur et les coauteurs le cas échéant conservent la propriété du droitd’auteur et des droits moraux qui protègent ce document. Ni la thèse ou lemémoire, ni des extraits substantiels de ce document, ne doivent êtreimprimés ou autrement reproduits sans l’autorisation de l’auteur.
Afin de se conformer à la Loi canadienne sur la protection desrenseignements personnels, quelques formulaires secondaires, coordonnéesou signatures intégrées au texte ont pu être enlevés de ce document. Bienque cela ait pu affecter la pagination, il n’y a aucun contenu manquant.
NOTICE
The author of this thesis or dissertation has granted a nonexclusive licenseallowing Université de Montréal to reproduce and publish the document, inpart or in whole, and in any format, solely for noncommercial educational andresearch purposes.
The author and co-authors if applicable retain copyright ownership and moralrights in this document. Neither the whole thesis or dissertation, norsubstantial extracts from it, may be printed or otherwise reproduced withoutthe author’s permission.
In compliance with the Canadian Privacy Act some supporting forms, contactinformation or signatures may have been removed from the document. Whilethis may affect the document page count, it does not represent any loss cfcontent from the document.
Université de Montréal
Faculté des études supérieures
Ce mémoire intitulé
Détection des gènes de résistance aux agents antimicrobiens chez des souches
d’Escherichia cou entérotoxinogènes et extraintestinales
présenté par
CHRISIINE MAYNARD
a été évalué par un jury composé des personnes suivantes
Sylvain Quessy, président-rapporteur
Josée Harel, directrice de recherche
Serge Larivière, codirecteur de recherche
Patrick Boerlin, membre du jury
111
Résumé
Dans le monde entier, une augmentation préoccupante de la prévalence de la résistance aux
agents antimicrobiens chez les bactéries est remarquée. L’utilisation des agents
antimicrobiens semble conduire vers une inévitable émergence de souches multirésistantes.
Escherichia cou est une bactérie qui a la capacité de se multiplier chez différents animaux à
sang chaud, y compris l’humain. Elle peut résider dans la flore commensale de son hôte mais
peut aussi provoquer des infections intestinales ou extraintestinales. Puisque cette bactérie
peut être retrouvée chez plusieurs espèces d’animaux, l’étude de celle-ci peut nous informer
des pressions sélectives et du processus d’évolution des bactéries qui existent en général.
Nous pouvons donc caractériser la bactérie E. cou comme étant une bactérie sentinelle.
Dans notre étude, nous avons voulu, dans un premier temps, vérifier comment la résistance
aux agents antimicrobiens codée par des gènes acquis retrouvés chez des souches appartenant
à un même groupe évoluait dans une période de temps donnée. Pour ce faire, un total de 11$
souches E. cou entérotoxinogènes (ETEC) appartenant au sérogroupe 0149: K91, isolées de
porcelets atteints d’une diarrhée postsevrage dans la province de Québec, entre 197$ et 2000,
et résistantes à au moins un agent antimicrobien, ont été caractérisées pour détecter leur profil
phénotypique et génotypique de résistance aux agents antimicrobiens.
Parmi les dix agents antimicrobiens testés, la majorité des souches ETEC était résistante à la
tétracycline et aux sulfamides, alors qu’aucune n’était résistante aux céphalosporines. De
plus, entre 197$ et 2000, une augmentation du nombre de souches ETEC résistantes au
chloramphénicol et de la multirésistance ont été observées. La distribution des gènes de
résistance pour les F3-lactamines, les aminoglycosides, le chloramphénicol, la tétracycline, la
triméthoprime et les sulfamides a été déterminée par hybridation sur colonies et ce, pour
toutes les souches ETEC. Des différences significatives dans la distribution des gènes de
résistance parmi ces souches, selon leur période d’isolement, ont été observées. En effet, la
distribution des gènes de résistance à la tétracycline (tet(A), tet(B), tet(C)), à la triméthoprime
(dhfrL dhfrV, dIfrXIII) et aux sulfamides (suiL sutil) a varié parmi les souches ETEC dans le
temps (197$-2000). Certains des gènes de résistance ciblés dans l’étude montraient certaines
associations entre eux. Parmi les souches ETEC, il a été démontré que les gènes tem et shv, de
même que les gènes tet(A) et tet(C) sont positivement associés entre eux de façon
iv
statistiquement significative. L’association qui peut exister entre les gènes est influencée par
la présence d’éléments génétiques mobiles. Ainsi, l’intégron de classe 1, a été détecté dans
63% des souches ETEC. Quatre régions variables de longueurs différentes d’intégron de
classe 1 ont été détectées par PCR et ont été caractérisées. Toutefois, la multirésistance
observée dans cette étude ne semble pas être complètement associée à cet intégron.
Dans une seconde étude, la distribution des gènes acquis de résistance aux agents
antimicrobiens a été analysée chez des souches E. cou appartenant au pathotype d’infections
extraintestinales (ExPEC), isolées de différents hôtes. Pour atteindre cet objectif, 109 souches
E. cou d’infections extraintestinales (ExPEC) résistantes à au moins un agent antimicrobien
testé, dont 39 souches provenaient de différents tissus d’animaux et de 70 souches isolées
d’infections urinaires d’humains, ont été caractérisées pour établir leur profil phénotypique et
génotypique de résistance aux agents antimicrobiens.
Parmi les souches ExPEC isolées d’animaux et d’humains, la résistance à l’ampicilline, à la
tétracycline et aux sulfamides était fréquente tout comme chez les ETEC. Toutefois, une
résistance aux céphalosporines n’a été détectée que chez les souches d’origine animale.
Plusieurs souches ExPEC démontraient une résistance à plus de trois agents antimicrobiens.
La distribution des gènes de résistance a été déterminée par hybridation sur colonies. Parmi
les souches ExPEC animales et humaines, la distribution des gènes de résistance est
différente. La répartition des gènes correspondant à une résistance à la tétracycline (tet(D)), au
chloramphénicol (catL catIIL foR) et à la triméthoprime (dhfrL dhfrV, dhfrVIJ dhfrXIII) est
différente selon l’origine de la souche (animale ou humaine). Parmi les souches ExPEC, les
mêmes associations positives retrouvées chez les souches ETEC ont pu être observées sauf
pour les gènes tem et shv. Environ le tiers des souches ExPEC isolées d’animaux (3 6%) et des
souches ExPEC isolées d’humains (3 1%) possédaient un intégron de classe 1. Quatre régions
variables de longueurs différentes, situées sur ces intégrons, ont été amplifiées par PCR parmi
les souches ExPEC et ont été caractérisées.
En conclusion, il a été démontré que parmi des souches E. cou d’origine porcine
entérotoxinogènes de sérogroupe 0149: K91 et des ExPEC provenant d’animaux et
d’humains, différents gènes de résistance peuvent être associés à un phénotype donné. En
effet, nos études démontrent que parmi les souches E. cou isolées à différentes périodes, les
V
gènes responsables de la résistance aux agents antimicrobiens ne sont pas les mêmes. De plus,
les gènes responsables de la résistance chez les ExPEC diffèrent selon la provenance de la
souche (animale ou humaine). La mobilité des gènes de résistance aux agents antimicrobiens
ainsi que les pressions sélectives exercées sur les bactéries, jouent un rôle dans
l’établissement des profils génétiques de la résistance bactérienne. Sans une analyse
génotypique de la résistance, la comparaison des profils de résistance entre deux bactéries
peut être erronée et amener à des conclusions incomplètes. Ces dernières conclusions
confirment l’importance de la caractérisation génotypique de la résistance des souches
bactériennes lors d’ études épidémiologiques.
Mots clés Escherichia cou entérotoxinogène! infection extraintestinale! intégron]
hybridation sur colonies! résistance! antibiotique.
vi
Summary
Around the world, an increase of the prevalence of antimicrobials resistance in bacteria is
observed. The use of antimicrobials in human and animals seems to lead to an inevitable
emergence of resistant strains. Escherichia cou is a bacterium which has the capacity to
multiply in different mammals, including humans. It can reside in the commensal fora of its
host but it can also cause intestinal or extraintestinal infections. Since this bacterium can be
isolated from different animal species, it can inform us on selective pressures and on the
evolution ofbacteria, in general. So, we can characterised E. cou like a sentinel bacterium.
In a first step, we wanted to evaluate the dyiiamic of the acquired resistance genes over time.
To do that, a total of 118 enterotoxigenic Escherichia cou (ETEC) 0149: K91 strains isolated
from pigs with postweaning diarrhoea collected ii the province of Quebec between 197$ and
2000, and presenting at Ieast one antimicrobial resistance, were characterised for their
phenotypic and genotypic antimicrobial resistance profile to ten antimicrobials.
The majority of the ETEC strains were resistant to tetracycline and to sulfonamides, whereas
resistance to cephalosporins vas not detected. Also, an increase in multiresistant strains
frequently was observed over time, particularly during the most recent period, 1995-2000.
The distribution of the resistance genes for 3-lactams, arninoglycosides, chloramphenicol,
tetracycline, trimethoprim and sulfonamides was assessed by colony hybridisation on ail
isolates. Significant differences in the distribution of tetracycline (tetA, tetB, tetC),
trimethoprim (dIfrL dhfrV, dhfrXIII) and sulfonamides (sulI, sulli) resistance genes were
observed among the ETEC strains according the their isolation period (197$ to 2000). Some
target resistance genes in this study showed associations between them. Among the ETEC
strains, a positive association of tein with shv as well as tet(A) with tet(C) was found. The
existence of associations between two genes could be influenced by the presence of mobile
genetic determinants in the strains. One of these determinants, class 1 integron, was detected
in 63% of the ETEC strains. Four variable regions with different lengths, situated on these
classl integrons, were amplified with PCR and were characterized.
In a second step, the distribution of antirnicrobial resistance acquired genes among E. cou
strains belonging to the extraintestinal pathogenic E. cou (ExPEC) serogroup, isolated from
vii
different hosts was evaluated. b reach this objective, 109 ExPEC strains resistant to at least
one antimicrobial tested, composed of 39 strains isolated from different infected animal
tissues and 70 strains isolated from human urine isolates, were characterised for their
phenotypic and genotypic antimicrobial resistance profile to ten antirnicrobials.
Among the animal and human ExPEC strains, the resistance to ampicillin, to tetracycline and
to sulfonamides were the most frequent resistance. However, resistance to cephalosporins was
detected only among the ExPEC strains of animal origin. Antimicrobial multiresistance was
common among our ExPEC strains. The distribution of the resistance genes was assessed by
colony hybridisation on ail isolates. Different genotypic profiles of the resistance genes were
established. In fact, the distribution of the genes corresponding to a resistance to tetracycline
(tet(D)), to chloramphenicol (catL catIIL foR) and to trimethoprim (dÏfrI, dhfrV, dÏfrVII,
dhfrXIII) was different between the animal and human isolates. In the ExPEC strains, the
same positive associations as with ETEC between genes were obsewed except for the
association of tem with shv. Thirty six percent of the animal ExPEC and 31% of the human
ExPEC strains had a class 1 integron. four variable regions with different lengths, situated on
these classi integrons, were ampÏffied by PCR and were characterised.
In conclusion, this study shows that among ETEC 0149: K91 porcine strains and ExPEC
strains isolated from animais and humans, the resistance genes are not static. Actually, our
studies demonstrated that among E. cou strains isolated from different time penods, the genes
responsible of the antimicrobial resistance were flot the same and changed. Also, according to
the isolation origin (animal or human), the genes responsible of the antimicrobial resistance in
ExPEC changed. These resuits reinforce the necessity to use genotypic resistance anaiysis in
future bacterial epidemiology studies.
Keywords : Enterotoxigenic Escherichia cou! extraintestinal infections! integron! colony
hybridisation! antimicrobial resistance! antibiotic.
viii
Table des matières
Identification du jury ii
Résumé iii
Summary vi
Table des matières viii
Liste des tableaux xiii
Liste des figures xv
Liste des sigles et abréviations xvi
Remerciements xx
SECTION I — Introduction I
Problématique 2
Recension de la littérature 4
1.0 L’utilisation des agents antimicrobiens et le phénomène de résistance bactérierm. .
2.0 Transfert horizontal de la résistance, devons-nous craindre l’animal ? 10
3.0 La bactérie Escherichia cou 14
4.0 L’intégron 17
5.0 La génétique de la résistance 20
6.0 Les agents antimicrobiens 24
6.] Les ,8-Ïactamines 24
6.2 Les aminogÏycosides 27
6.3 Les tétracyclines 31
6.4 Les phénicols 33
6.5 La triméthoprime 35
6.6 Les sufamides 38
7.0 Contrôle de la résistance antirnicrobienne 41
ix
SECTION II — Méthodologie et Résultats.42
Article I : Characterization and detection of antimicrobial resistance genes in enterotoxigenic
Escherichia cou 0149: K91 isolated from pigs over a 23-year period 43
Characterization and detection of antimicrobial resistance genes in enterotoxigenic
Escherichia cou 0149: K91 isolated from pigs over a 23-year period 44
Abstract 45
Introduction 46
Materials and methods 48
Bacterial strains and growth conditions 48
Antimicrobiat susceptibility testing 48
FCR prirners and ampÏfication 49
DNA sequencing 49
Cotony hybridization 50
Statistical methods 50
Resuits 51
Antimicrobial resistancephenotype characteristics 51
Distribution ofresistance genes in 0149: K91 ETEC isotates 51
(j) Beta-lactams 51
(ii) Aminoglycosides 52
(iii) Tetracycline 52
(iv) Phenicols 52
(y) Trimethoprim 52
(vi) Sulfonamides 53
Identification ofintegrons 53
Association between resistance genes 54
Discussion 55
Acknowledgments 59
References 60
X
Tables.64
Article II : Genetic analysis of antirnicrobial resistance among extraintestinal Escherichia
cou isolates of animal and human origin 72
Genetic analysis of antimicrobial resistance among extraintestinal Escherichia cvii
isolates of animal and human origin 73
Synopsis 74
Introduction 75
Materials and methods 77
Bacterial strains and growth conditions 77
Antimicrobial susceptibility testing 77
FCR primers and amphfication 77
DNA sequencing 78
Colony hybridization 78
StatisticaÏ rnethods 78
Resuits 80
Antimicrobial resistance phenotype characteristics 80
Distribution ofresistance genes 80
(j) Beta-lactams 80
(ii) Arninoglycosides 81
(iii) Tetracychne 81
(iv) Phenicols 81
(y) Trimethoprim 82
(vi) Sulfonamides 82
Identification and characterisation of integrons 82
Distribution ofvirulence genes 83
Association betveen resistance and virulence genes 83
Phylogenetic grouping resuits 83
xi
Discussion . $5
AcknowÏedgrnents .90
References 91
Tables 9$
SECTION III — Discussion 106
Discussion 107
Étude sur les isolats Escheric!tia cou entérotoxinogènes 107
1.0 Profil génotypique de la résistance 10$
1.0.1 Sélection de la résistance 108
1.0.2 Émergence de la inutirésistance 110
1.1 Profil génotypique de la résistance 111
1.1.1 Diversité des gènes 111
1.1.2 Mobilité des gènes de résistance 114
1.2 Association des gènes de résistance et co-sélection 116
1.3 Changements factoriels et diversité des profils 117
1.4 Expériences et analyses à venir 119
Étude sur les souches Escitericitia cou pathogènes extraintestinales d’origine animale et
humaine 120
2.0 Profil génotypique de la résistance 120
2.0.1 Vision d’ensemble 120
2.0.2 Hoînogénéité dans les profils de ,nttÏtirésistance 121
2.1 Profil génotypique de résistance 122
2.1.1 Diversité des gènes 122
2.1.2 Mobilité des gènes 123
2.2 Association et co-sélection des gènes de résistance 124
2.3 Virulence et phylogénie des souches 124
xii
2.3.1 Profil de virulence.125
2.3.2 Association virulence et résistance 126
Conclusion 127
SECTION IV— Bibliographie 129
SECTION V—Annexes xxi
Annexe 1 : Évolution de la multirésistance chez les souches Escherichia cou
entérotoxinogènes xxii
Annexe 2 : Analyse statistique du du-carré entre différents gènes de résistance et
l’intégron de classe 1 retrouvés parmi les souches ETEC isolées entre 1978 et 1989. xxiii
Annexe 3 : Analyse statistique du chi-cané entre différents gènes de résistance et
l’intégron de classe 1 retrouvés parmi les souches ETEC isolées entre 1990 et 2000 ...xxiv
Annexe 4: Distribution du phénotype de résistance aux agents antimicrobiens selon l’âge
des porcs xxv
Annexe 5 : Distribution de l’âge des porcs selon l’année d’isolement de la souche .... xxvi
xiii
Liste des tableaux
SECTION I — Iittroductioit I
Recension de la littérature 4
Tableau I: Agents antimicrobiens employés au Canada en médecine
vétérinaire et chez les humains selon le type d’utilisation 13
Tableau II: Mécanismes d ‘action des déterminants génétiques de résistance
22
SECTION II — Méthodologie et Résultats 42
Article I : Characterization and detection of antimicrobial resistance genes in enterotoxigdnic
Escherichia cou 0149: K91 isolated from pigs over a 23-year period 43
Table 1: PCR primers usedfor antimicrobiat resistance gene and class 1
integron ampflfications 64
Table 2 : Trends in resistance ofETEC 0149 K9] isotates with time as
determined by the disc dffusion method 66
Table 3 Distribution ofantimicrobial resistance genes detected in ETEC
0149 K9] isolates according to tue isolation periods 67
Table 4 Distribution ofclass 1 integron among the ETEC 0149 K91
isolates 70
Table 5 Association between the various antimicrobial resistance genes
and class 1 integron among ETEC 0149 K9] isolates 71
Article II : Genetic analysis of antimicrobial resistance among extraintestinal Escherichia
coli isolates of animal and human origin 72
Table I: PCI? primers usedfor antimicrobial resistance gene amplifications
98
Table II: Number ofresistant animal and human isoÏates determined by disc
dffitsion ,nethod 100
xiv
Table III: Animal and human isolates which gave a positive hybridisation
with the resistance and virulence tested probes and had the associated
resistancephenorype 101
Table IV: Distribution ofclass] integron among the animal and human E.
cou strains 103
Table V : Association between antimicrobial resistance genes, class 1
integron and virulence genes among the ExPEC strains isoÏated from
animais 104
Table VI : Association between antimicrobial resistance genes, class J
integron and virulence genes among the ExPEC strains isolated froin
humans 105
SECTION V-Annexes xxi
Tableau III. Association des gènes de résistance et e 1 ‘intégron de classe 1
parmi les souches ETEC isolées entre 1978 et 1989 xxiii
Tableau IV. Association des gènes de résistance et e 1 ‘intégron de classe 1
parmi les souches ETEC isolées entre 1990 et 2000 xxiv
xv
Liste des figures
SECTION I — Introduction 1
Recension de la littérature 4
Figure] : Structure générale de 1 ‘intégron de classe] 19
SECTION V-Annexes xxi
Figure 2 : Évohttion ascendante de la mttltirésistance chez les souches ETEC
isolées de porcs atteints d ‘ttne diarrhée postsevrage entre les années 1978 et
2000 xxii
Figttre 3 Pourcentage de souches ETEC selon le nombre de résistances aux
agents antimicrobiens testés en fonction de Ï ‘âge des porcs d ‘où proviennent
les souches xxv
Figure 4: Pourcentage de souches ETEC selon l’âge des porcs d’où
proviennent tes souches en fonction de la période d ‘isolement de celles-ci
xxvi
xvi
Liste des sigles et abréviations
AAC Acétyltransférase
ADN (DNA) Acide désoxyribonucléique (desoxyribonucleic
acid)
ACIA Agence canadienne d’inspection des aliments
afa afimbrial adhesin
AIVIP Ampicilline
AMK Amoxicilline
AMX Amoxicilline-acide clavulanique
ANT Adénylyltransférase
APH Phosphotransférase
APR Apramycine
ARN Acide ribonucléique
ATM Aztreonam
CAR Carbenicilline
CAT Chloramphénicol acétyltransférase
CE Communauté européenne
CHL Chloramphénicol
CMI Concentration minimale inhibitrice
CR0 Ceftriaxone
CS Conserved sequences
CTX Cefotaxime
DHfR Dihydrofolate réductase
DHP$ Dihydroptérate synthase
DPT Direction des productions thérapeutiques
DMV Direction des médicaments vétérinaires
dNTP deoxyribonucleoside triphosphate
E. aerogenes Enterobacter aerogenes
E. cou Escherichia cou
EHEC Enterohemorrhagic Escherichia cou
Els Extraintestinal infections
EPEC Enteropathogenic Escherichia cou
xvii
ESBL Extended-spectrum 3-lactamase
ETEC Enterotoxigenic Escherichia cou
ExPEC Extraintestinal pathogenic E. cou
ERV Entérocoques résistants à la vancomycine
FDA Food and drug administration
FFL Florfénicol
FPPQ Fédération des producteurs de porcs du Québec
GCG Genetics computer group
GEN Gentamicine
GREMIP Groupe de recherche sur les maladies du porc
1$ Insertion sequences
KAN Kanamycine
K. oxytoca KlebsieÏÏa oxytoca
LB Luria Bertani
LCR Ligase chain reaction
MH Mueller-Hinton
MIN Minocycline
NCCLS National committee for clinical laboratory standards
NEO Néomycine
NSERC Natural sciences and engineering research council of
Canada
ORF Open reading frame
PABA para-aminobenzoic acid
pap pyelonephritis-associated pili
PBP Penicillin binding protein
PCR Polymerase chain reaction
PIP Piperacilline
RAPD Random amplification of polymorphic DNA
RNSM Recueil des notices sur les substances médicatrices
du Canada
RV (VR) Région variable (variable region)
TET Tétracycline
TIC Ticarcilline
xviii
11M Ticarcilline-acide clavulanique
TMP Triméthoprime
TOB Tobramycine
S. enterica Saïrnonella enterica
sfa S fimbrial adhesin
S. rimosus Streptoniyces rimosus
SRRP Syndrome reproductif et respiratoire porcin
STEC Shiga toxin-producing E. cou
SUL Sulfamides
UTI Urinary tract infection
155 Tryptic soy broth
TSA Tiyptic soy agar
Unités de mesure
C celcius
g gramme
kb Kilobase
kg Kilogramme
min minute
ml millilitre
mM millimolaire
pb (bp) Paire de bases (base pairs)
pmol picomole
sec seconde
microgramme
.tl microlitre
micromolaire
U unité
v/v volume/volume
w / y weight / volume
Symboles
bêta
lambda
degrés
± plus ou moins
plus grand ou égal à••
plus petit ou égal à••
xix
xx
Remerciements
Je tiens à remercier sincèrement tous ceux et celles qui ont rendu ce projet de maîtrise
réalisable, et plus particulièrement:
Serge Larivière, d’avoir cru en ce projet dès le départ et su me montrer l’ampleur de celui-ci
Josée Harel, pour sa grande générosité, pour ses connaissances et surtout d’avoir su pousser
mes limites
Sadjia Bekal, pour sa patience, son tact et de m’avoir appris la détermination;
Madeleine fortin, pour son dévouement et sa gentillesse;
les membres du GREMIP, chercheurs, collèges et ami(e)s, pour le partage de vos idées, de
vos connaissances et pour tous les bons moments passés ensemble;
mes amies qui ne m’auront pas tellement vu pendant ces dernières années et qui continuent à
m’offrir leur amitié;
Julie Parenteau, pour avoir posé son regard sur mes mots parfois défectueux;
Michel et francine, pour leur amour, leur patience et de m’avoir endurer pendant ces deux
années (et plus !)
Dominik, pour sa compréhension et son soutien;
la Fédération des producteurs de porcs du Québec (FPPQ), pour avoir cm bon d’investir dans
ce projet;
et finalement, je remercie le CRSNG et le réseau canadien de recherche sur les bactéries
pathogènes du porc pour le financement de ce projet de maîtrise.
SECTION I - Introduction
2
Problématique
Escherjchia cou (E. cou) est une cause importante d’infections chez l’animal et chez l’humain
dans le monde entier (Bopp, C. A. et aÏ. 2003). Cette bactérie peut être divisée en trois
groupes : souches commensales, souches pathogènes intestinales et souches causant des
pathologies extraintestinales. Les deux derniers groupes sont ceux qui retiennent plus
particulièrement l’attention à cause de leur pouvoir pathogène. Les souches pathogènes
intestinales peuvent être à leur tour divisées en plusieurs groupes dont quatre majeurs : E. cou
entérotoxinogène (ETEC), entéropathogénique (EPEC), entéroinvasive (EIEC),
entéroaggrégative (EAggEC).
Les souches ETEC sont des agents pathogènes importants du porc causant des diarrhées chez
les nouveau-nés et chez les porcelets en postsevrage. Les ETEC de sérogroupe 0149 : K9 1
sont isolées de plus en plus fréquemment. D’ailleurs, les souches ETEC appartenant à ce
sérogroupe sont isolées dans le monde entier et représente le sérogroupe prédominant associé
aux diarrhées chez le porc (Wray, C. et al. 1993; Garabal, J. I. et al. 1996; Wada, Y. et aï.
1996; Nagy, B. et al. 1990; Wieler, L. H. et al. 2001; Amezcua, R. et al. 2002). Avec
l’augmentation mondiale de la résistance aux agents antimicrobiens chez E. cou, le traitement
des diarrhées en postsevrage devient très difficile. Les souches ETEC développent des
résistances aux agents antimicrobiens qui sont employés présentement, peu importe le mode
d’action ou la voie d’administration de ceux-ci (Amezcua, R. et al. 2002; Bischoff, K. M. et
aÏ. 2002).
Par ailleur, il existe des souches E. cou causant des infections extraintestinales appelées
ExPEC. Les ExPEC peuvent causer des infections à plusieurs sites anatomiques différents
chez l’animal, mais aussi chez l’humain (Russo, T. A. and Jolmson, J. R. 2000). L’usage des
agents antimicrobiens dans le traitement d’infections extraintestinales causées par E. cou est
un outil important dans la guérison de celles-ci. Cependant, la résistance aux agents
antimicrobiens est répandue et touche gravement la médecine vétérinaire et celle des humains
(Aubry-Damon, H. and Courvalin, P. 1999; Monroe, S. and Polk, R. 2000; Teshager, T. et al.
2000). Une association existe entre l’utilisation d’agents antimicrobiens et le développement
3
de souches résistante (Schwarz, S. and Chaslus-Dancla, E. 2001; Aarestrup, f. M. and
Wegener, H. C. 1999; Guillemot, D. 1999). La plupart des classes d’agents antimicrobiens
utilisées en médecine pour les humains sont aussi employées en médecine vétérinaire. Chez
les animaux, les agents antimicrobiens sont non seulement utilisés pour contrôler les
infections, mais aussi pour stimuler leur croissance (Santé Canada, 2002). Conséquemment,
nous nous inquiétons du fait que l’utilisation des agents antimicrobiens chez les animaux
puisse contribuer directement ou indirectement à l’émergence de bactéries pathogènes
résistantes aux agents antimicrobiens chez les humains. Ceci est toutefois difficile à
démontrer. Afin d’étayer ce lien, les études de caractérisation de la résistance génotypique des
souches pourraient être utiles. Toutefois, la majorité des études portant sur la résistance aux
agents antimicrobiens se concentrent plutôt sur l’augmentation de la résistance phénotypique
et très peu portent sur la caractérisation de cette résistance.
Le premier objectif de notre étude consiste à vérifier si la résistance aux agents antimicrobiens
des souches appartenant à un même groupe évolue dans le temps. Pour ce faire, nous avons
identifié certains des gènes de résistance aux agents antimicrobiens causant la résistance
phénotypique retrouvée chez des souches ETEC isolées de porcelets atteints d’une diarrhée en
postsevrage. Ces souches de sérogroupe 0149 K91 ont été isolées à des périodes différentes,
entre 197$ et 2000.
Le deuxième objectif de notre étude consiste à vérifier si les profils de résistance aux agents
antimicrobiens des souches appartenant à un même groupe, mais isolées d’hôtes différents,
sont différents. Afin d’atteindre cet objectif, nous avons déterminer la présence de certains
gènes de résistance aux agents antimicrobiens chez des souches ExPEC isolées d’infections
extraintestinales d’animaux et d’humains.
4
Recension de la littérature
1.0 L’utilisation des agents antimicrobiens et le phénomène de résistance bactérienne
Les antibiotiques sont décrits comme des substances produites par des organismes vivants
qui sont, à faibles concentrations, capables d’inhiber la croissance d’autres organismes.
Toutefois, on a souvent tendance à inclure dans cette définition des substances synthétiques
utilisées comme médicaments anti-infectieux comme, par exemple, les sulfamides. Dans ce
cas, il est préférable d’utiliser un terme plus général, celui d’agents antimicrobiens. Les
agents antimicrobiens sont surtout employés pour combattre les infections chez l’humain,
les animaux et plantes, mais sont également utilisés pour préserver certains composés
biologiques ainsi que pour favoriser la croissance de certains animaux.
Dans le monde entier, on remarque qu’il y a une augmentation de la prévalence de la
résistance bactérienne aux agents antimicrobiens. Il est généralement accepté que le facteur
majeur de l’augmentation de la résistance est dû à l’augmentation de l’utilisation des agents
antimicrobiens (Aarestrup, f. M. and Wegener, H. C. 1999; Health Canada 2002; Alekshun,
M. N. and Levy, S. B. 2000). L’utilisation des agents antimicrobiens semble conduire vers
une émergence de gènes de résistance et de souches résistantes à cause de leur sélection. Cette
situation est applicable à l’usage vétérinaire et humain des agents antimicrobiens.
Environ la moitié des agents antimicrobiens fabriqués aux États-Unis seraient distribués aux
animaux (van den Bogaard, A. E. and Stobberingh, E. E. 2000). Les agents antimicrobiens
employés en indication vétérinaire ne sont pas seulement utilisés en traitement ou en
prévention d’infections bactériennes, ils sont aussi administrés dans les aliments donnés à
l’animal afin de promouvoir leur croissance et d’augmenter la conversion alimentaire (van
den Bogaard, A. E. and Stobberingh, E. E. 2000).
Plusieurs études rétrospectives et prospectives ont étudié l’émergence de la résistance
bactérienne chez l’animal à la suite de l’utilisation d’agents antimicrobiens. Plusieurs de ces
études montrent qu’après l’introduction d’un agent antimicrobien dans la pratique vétérinaire,
5
la résistance des bactéries pathogènes comme des bactéries de la flore intestinale, augmente.
Toutefois, peu d’études se concentrent sur la résistance aux agents antimicrobiens qui sont
utilisés comme facteur de croissance chez les animaux. En ce qui concerne la bactérie
Escherichia cou, une augmentation significative de souches résistantes au carbadox, un agent
antimicrobien approuvé au Canada en 1970, a été notée après que cet agent ait été employé
comme facteur de croissance dans l’élevage du porc (Ohmae, K. et al. 1981). Toutes les
souches E. cou résistantes au carbadox possédaient le même plasmide transférable causant la
résistance à cet agent. Le carbadox n’a pas été utilisé dans l’élevage des poulets et aucune
résistance à celui-ci n’a été détectée chez les souches E. cou isolées des poulets de la même
région. Ohmae et al. ont noté une augmentation du pourcentage de souches E. cou résistantes
au carbadox de 37 à 61% après l’introduction de celui-ci. Cependant, le carbadox n’est pas
seulement employé comme facteur de croissance chez les porcs, mais aussi en prévention à la
dysenterie porcine et, à plus fortes doses, pour le traitement des salmonelloses.
L’intérêt des agents antimicrobiens utilisés comme facteurs de croissance chez les animaux a
augmenté après l’émergence d’entérocoques résistants à la vancomycine (ERV) chez
l’humain. En Europe, l’avoparcin, un agent antimicrobien apparenté à la vancomycine, a été
rapidement pointé du doigt comme étant la cause de cette résistance. Dans les pays où cet
agent antimicrobien a servi de facteur de croissance, les ERV n’ont pas été seulement isolés
d’animaux ayant reçu cet agent, mais aussi chez des animaux et des humains en bonne santé
et n’ayant jamais reçu de l’avoparcin (Bates, J. et al. 1994).
La Suède interdit l’usage des facteurs de croissance dans la nourriture des animaux depuis
1986. La prévalence de la résistance aux agents antimicrobiens utilisés comme facteurs de
croissance en 1997, était beaucoup plus faible chez les porcs de Suède comparativement aux
porcs de la Suisse (van den Bogaard, A. E. and Stobberingh, E. E. 2000). De plus, en Suède
tout comme aux États-Unis, où l’avoparcin n’a jamais été employé, un faible taux de ERV
résistants à cet agent antimicrobien, dans les isolats fécaux d’animaux sains et d’humains à
l’extérieur des hôpitaux, a été détecté (Coque, T. M. et al. 1996).
6
Par contre, les restrictions quant à l’utilisation des agents antimicrobiens comme facteur de
croissance chez les animaux destinés à la consommation humaine, ne semblent pas permettre
de contrôler entièrement la résistance microbienne. En fait, il semblerait que la résistance à
certains agents antimicrobiens persiste après l’arrêt de l’usage vétérinaire de ces agents.
Certains pays européens ont interdit, vers 1995, l’utilisation de la vancomycine (avoparcin),
l’avilamycine, l’érythromycine et la virginiamycine comme facteur de croissance chez les
animaux (Aarestrup, F. M. et al. 2001). Cinq années après leur interdiction, une baisse
significative de la résistance à ceux-ci chez les entérocoques isolés des porcs, sauf dans le cas
de la résistance à la vancomycine, a été observé (Aarestrup, F. M. et al. 2001; Boerlin, P. et
al. 2001). La persistance de la résistance à la vancomycine (avoparcin) chez Enterococcus
faecium pourrait être due à l’utilisation par ces pays de la tylosine, un macrolide, jusqu’en
1999. Le mécanisme de résistance offrant une résistance à la vancomycine, a peut-être été
favorisé par l’utilisation de la tytosine. Par contre, maintenant que tous les agents
antimicrobiens employés comme facteurs de croissance ont été bannis par plusieurs pays de
l’Europe, il ne semble pas que la fréquence de souches d’ERV ait diminué (Heuer, O. E. et al.
2002).
Nous savons que le développement de la résistance aux agents antimicrobiens est affecté par
trois facteurs (1) le volume d’agents utilisés, (2) le taux de formation de mutants résistants et
(3) le coût biologique de la résistance et ses compensations génétiques (Bjorkman, J. and
Andersson, D. I. 2000). Le fait que l’arrêt de l’utilisation des agents antimicrobiens n’amène
pas automatiquement une régression de la résistance à ceux-ci, peut être l’effet d’une
compensation due à des mutations au niveau soit du chromosome bactérien ou du plasmide
qui transporte la résistance (Bjorkman, J. and Andersson, D. 1. 2000). Ces mutations
permettent aux bactéries de contourner les coûts biologiques que la résistance peut infliger.
Par exemple, il a été démontré que certains mutants de SalmoneÏla typhiinuritun, qui étaient
résistants à la streptomycine, étaient moins virulents que la souche sauvage non résistante
(Maisnier-Patin, S. et al. 2002). Alors, dans des conditions non sélectives (sans
streptomycine), le mutant résistant était défavorisé comparativement à la souche sauvage.
Toutefois, certains de ces mutants avaient retrouvé leur virulence tout en conservant la
résistance à la streptomycine. Ces mutants possédaient une substitution compensatoire d’un
7
acide aminé dans un de leurs gènes chromosomiques ips. Ceci démontre que la résistance à
un agent antimicrobien créée par une mutation chromosomique peut être coûteuse à la
bactérie qui la possède, mais que celle-ci peut compenser ces effets néfastes par une autre
mutation.
Nous remarquons ce phénomène aussi parmi tes bactéries qui sont résistantes aux agents
antimicrobiens par l’acquisition d’un plasmide. En effet, certains plasmides amènent des
inconvénients à la bactérie qui le possède. C’est le cas du plasmide pBR322 qui porte le gène
tet(C) et qui transmet la résistance à la tétracycline. L’expression du plasmide réduit l’activité
de l’enzyme ATPase de la bactérie, ce qui ralenti considérablement la vitesse de croissance de
la bactérie (Valenzuela, M. S. et aÏ. 1996). En absence de tétracycline, la bactérie peut se
débarrasser du plasmide ou, alternativement, le conserver et créer une mutation compensatoire
afin de n’être pas désavantagée par le port du plasmide.
Certaines bactéries ou mutants ont la capacité de créer un taux élevé de mutations et sont donc
plus sujettes à posséder des mutations compensatoires aux inconvénients associés à la
résistance aux agents antimicrobiens. Seul un faible pourcentage des bactéries E. cou et S.
ente,-ica possèdent cette caractéristique. Cependant, un tiers des bactéries Pseudomonas
aeruginosa possède un phénotype «mutateur» (Bjorkman, J. and Andersson, D. I. 2000).
L’approbation de tous les agents antimicrobiens, incluant les facteurs de croissance, est
réglementée afin de réduire leur utilisation. Santé Canada réglemente présentement la vente
des médicaments au Canada grâce à la Loi sur tes aliments et drogues et de son Règlement
ainsi que de la Loi réglementant certaines drogues et autres substances. Pour ce qui est des
médicaments pour usage humain, c’est la Direction des produits thérapeutiques (DPT) qui
administre principalement ces lois. La Direction des médicaments vétérinaires (DMV),
anciennement le Bureau des médicaments vétérinaires, administre ces lois pour les
médicaments à usage vétérinaire, dont les agents antimicrobiens pour les animaux destinés à
l’alimentation. La DMV est responsable des questions de sécurité pour l’alimentation
humaine concernant les médicaments à usage vétérinaire. Actuellement, il n’existe pas à la
DMV de méthodes et de critères particuliers pour évaluer la sécurité pour la santé humaine
8
des médicaments à usage vétérinaire, en ce qui a trait à la résistance aux agents
antimicrobiens (Santé Canada. http://www.hc-sc.gc.ca/vetdrugs
medsvet/amr/pdt7f_chapter4.pdf). Cela s’appliqtie aussi à la sécurité pour les animaux.
Les médicaments à usage vétérinaire sont classés en groupes, d’après une approche basée sur
la gestion du risque:
1. Les médicaments contrôlés sont utilisés pour une thérapie donnée sous le strict
contrôle d’un vétérinaire. Ce groupe de médicaments comprend des produits comme
les stimulants, les anesthésiques et les sédatifs;
2. Les médicaments à usage vétérinaire non inscrits sont ceux qui sont vendus sans
ordonnance, comme l’aspirine;
3. Les médicaments destinés à l’utilisation vétérinaire qui demandent l’ordonnance d’un
pharmacien, d’un praticien (c’est-à-dire, un vétérinaire) ou d’un fabricant titulaire
d’une licence;
4. Les médicaments inscrits qui peuvent se vendre sans ordonnance et sont étiquetés
ainsi.
5. Les aliments médicamenteux : le Recueil des notices sur les substances médicatrices
du Canada (RNSM) donne la liste des ingrédients médicamenteux (y compris les
agents antimicrobiens) approuvés par Santé Canada pour une utilisation dans les
aliments.
De plus, tous les produits médicamenteux n’ayant que des allégations thérapeutiques ne
peuvent servir de stimulateur de croissance, même avec une ordonnance vétérinaire. Les
médicaments, y compris les stimulateurs de croissance, sont approuvés pour une utilisation
dans les aliments et inclus dans le RNSM sur la base d’allégations faites par les fabricants du
médicament. Une allégation représente une utilisation spécifique, un taux d’utilisation et une
formulation du produit pour une substance médicatrice particulière.
Au Canada, les provinces possèdent leur propre organisme de contrôle, ainsi que le droit de
réglementer plus strictement la vente de médicaments, y compris les agents antimicrobiens,
une fois ceux-ci approuvés au niveau fédéral. Certaines provinces ont promulgué leur propre
9
législation. Au Québec, les médicaments à usage vétérinaire sont réglementés par la Loi sur la
pharmacie, la Loi sur les médecins vétérinaires et la Loi sur la protection sanitaire, la
sécurité et le bien-être des animaux. Cette province limite la vente de médicaments à usage
vétérinaire aux pharmaciens et aux médecins vétérinaires. La réglementation concernant les
modalités et les conditions de la vente de médicaments contient cinq annexes. Les trois
premières fournissent la liste des médicaments pour les humains et les deux autres, la liste des
médicaments pour les animaux. Il existe des permis pouvant être émis aux personnes qui
fabriquent, distribuent et vendent des pré mélanges ou des aliments médicamenteux. Un
détenteur de ce permis doit obtenir et détenir une ordonnance vétérinaire pour vendre les
aliments médicamenteux. Un producteur peut préparer des aliments médicamenteux pour ses
propres animaux sans détenir un permis, en autant qu’il ne prépare pas plus d’un kilogramme
ou d’un litre d’aliment médicamenteux.
10
2.0 Transfert horizontal de la résistance, devons-nous craindre l’animal?
Présentement, on remarque que l’utilisation thérapeutique et sub-thérapeutique des agents
antirnicrobiens chez l’animal suscitent l’intérêt public et scientifique. Ceci provient
probablement de l’accroissement de la multirésistance chez les bactéries pathogènes à
caractère zoonotique (Bolton, L. F. et al. 1999). L’utilisation d’agents antimicrobiens amène
une sélection de la résistance chez les souches pathogènes et celles de la flore endogène chez
l’animal et l’humain. Les bactéries résistantes isolées chez l’animal peuvent rejoindre la
population humaine par contact direct, mais aussi par la voie des aliments d’origine animale.
Ces bactéries peuvent coloniser l’humain ou transférer leurs gènes de résistance à d’autres
bactéries de la flore endogène de l’humain (van den Bogaard, A. E. and Stobberingh, E. E.
2000).
Des bactéries colonisant l’intestin et appartenant à l’espèce E. cou, nonpathogènes et
résistantes aux agents antimicrobiens, représentent probablement un important réservoir de
gènes de résistance (Osterblad, M. et al. 2000). D’une part, les souches E. cou résistantes et
d’origine animale peuvent coloniser l’intestin de l’humain, et ce, pour au moins un certain
temps (Marshall, B. et aÏ. 1990). D’autre part, l’habileté des bactéries à transférer de
nouveaux gènes de résistance entre elles par la conjugaison, par exemple, contribue
significativement à l’émergence de souches résistantes
Il a été démontré que les traitements aux agents antimicrobiens ciblés à un individu peuvent
influencer la flore endogène d’un autre individu côtoyant celui-ci ou étant relativement près
physiquement de celui-ci. Une étude réalisée par Oppegaard et al (Oppegaard, H. et al. 2001)
en Norvège a montré que des souches E. cou isolées de vaches et d’humains ayant côtoyé ces
vaches montraient des profils de résistance semblables. Cependant, après le sérotypage de ces
souches, les auteurs en sont venus à la conclusion que la résistance avait été disséminée par
des événements de transferts horizontaux des gènes de résistance plutôt que par des transferts
d’isolats résistants.
11
Il est difficile de prouver le transfert de la résistance entre l’animal et l’humain, car les agents
antimicrobiens employés chez l’animal sont souvent utilisés chez l’humain et vise-versa
(Tableau I). Toutefois, dans les cas où l’agent antimicrobien est utilisé seulement chez
l’animal, certaines évidences d’un transfert ont été démontrées. Après l’introduction de la
streptothricine comme facteur de croissance durant les années 80 dans les élevages
d’animaux, une émergence de souches E. cou possédant un plasmide qui transmet la
résistance à la streptothricine a été notée. Ce plasmide a été isolé de souches provenant de
porcs, d’éleveurs de porcs et de leur famille (Hummel, R. et cd. 1986). De plus, de telles
observations ont aussi été notées après l’introduction de l’apramycine dans la médecine
vétérinaire. Un transfert direct de la résistance à l’apramycine de souches E. cou isolées de
porcs à d’autres isolées d’humains a été observé et proviendrait de la présence d’un gène de
résistance situé sur un plasmide (Hunter, J. E. et al. 1994; Hunter, J. E. et aÏ. 1993; Johnson,
A. P. et aÏ. 1994). En fait, puisque les pressions sélectives sont plus élevées chez l’animal que
chez l’humain, à cause de l’usage d’agents antimicrobiens fréquent, on suggère que la
résistance développée par les bactéries entériques apparaît premièrement chez l’animal et que,
par la suite, cette résistance est transmise aux humains (van den Bogaard, A. E. 1997).
Comme la plupart des gènes de résistance provoquent une résistance croisée à d’autres agents,
les plasmides ou les gènes de résistance peuvent être sélectionnés par l’usage d’autres agents
antimicrobiens utilisés chez l’humain.
Les agents antimicrobiens qu’on emploie afin de traiter les infections représentent de loin la
plus grande proportion d’agents utilisés chez l’animal. On estime que, même s’ils comptent
seulement 15% du total des agents antimicrobiens administrés, les agents antimicrobiens
employés comme facteurs de croissance, compte pour environ 50% (kg) des molécules
actives (comprenant les agents antimicrobiens, médicaments, vitamines, etc.) administrées à
l’animal (van den Bogaard, A. E. 1997). Une étude hollandaise a démontré que la quantité
d’agents antimicrobiens utilisée chez l’animal destiné à la consommation humaine équivaut à
la quantité d’agents antimicrobiens utilisée chez l’humain (van den Bogaard, A. E. 1997).
Selon certains auteurs, il demeure illogique de continuer à différencier les agents
antimicrobiens exploités comme facteurs de croissance à ceux destinés aux traitements
d’infections en médecine vétérinaire, puisque que tous les deux conduisent à la résistance
12
bactérienne. Il serait plus juste de penser que la diminution de l’administration d’agents
antimicrobiens, peu importe le type d’utilisation, diminuerait la résistance (van den Bogaard,
A. E. 1997).
En fait, tout usage d’agents antimicrobiens favorise l’émergence de bactéries résistantes aux
agents antimicrobiens. La question est de savoir jusqu’à quel point l’utilisation d’agents
antimicrobiens chez les animaux est responsable des problèmes de résistance antimicrobienne
chez l’humain.
13
Tableau I. Agents antimicrobiens employés au Canada en médecine vétérinaire et chez les
humains selon le type d’utilisation.
Approuvés pour différentes utilisations selon les dïfférentesespèces animalesa
Prévention,Classe d agent . Facteur de TherapieTherapie prophylaxie eUouantimicrobien croissance humainecontrole
Ampicilline B, P, C, Ch Non Non Oui
Ceftiofur B, P, C Non Non Non
Cefotaxime Non Non Non Oui
Gentamicine B, A, P, C, Ch Non A Oui
Néomycine B, C, P B B, P, C Oui
Kanamycine Non Non Non Oui
Chloramphénicol C, Ch Non Non Qujb
Florténicol B Non Non Non
Chlortétracycline B, A, P B, A, P B, A, P Qujb
Oxytétracycline B, A, P B, A, P B, A, P Oui
TétracyclineB A P C, Ch Non A Ouibhydrochloride
Doxycycline Ch Non Non Oui
Sulfadiazine B, P, C, Ch Non Non Oui
Sulfadiméthoxine B, P, C, Ch Non Non Oui’
Sulfaguanidine B, P, C, Ch Non B, P Oui
Sulfaméthazine B, A, P, C, Ch B, A, P B, P Non
Triméthoprime B, P, C, Ch Non Non Oui
B, Bovin; P, Porcin; A, Aviaire; C, Chien; Ch, Chat.b Rarement utilisé.
14
3.0 La bactérie Escherichia cou
La bactérie Escherichia cou a été décrite par Theodor Escherich en 1919 comme étant un
bâtonnet à Gram-négatif anaérobie facultatif et non-sporulant de la famille des
Enterobacteriaceae. Cette bactérie constitue une cause d’infection importante chez l’animal et
l’humain [Bopp, 2003 #322].
Les souches pathogènes intestinales E. cou causent des diarrhées chez des porcelets nouveau-
nés. Les souches responsables de ces diarrhées sont principalement des souches E. cou
entérotoxinogènes (ETEC) appartenant au sérogroupe 0149: K91 qui colonisent l’intestin
grêle et produisent des entérotoxines agissant sur les entérocytes de l’animal [Bertschinger,
1999 #227]. Les signes cliniques d’une infection causée par les ETEC peuvent être observés
quelques heures après la naissance ainsi que quelques jours après le sevrage des porcelets. Il
en résulte, dans la plupart des cas, une hypersécrétion d’eau et d’électrolytes ainsi qu’une
diminution de l’absorption.
Le traitement d’une infection aux ETEC nécessite habituellement l’utilisation d’agents
antimicrobiens de large spectre. Toutefois, une augmentation d’environ 20% du nombre de
souches ETEC multirésistantes, isolées de porcs atteints d’une diarrhée postsevrage, a été noté
entre 1980 et 2000 (Amezcua, R et al. 2002; Bertchinger, H. U. and fairbrother, J. M. 1999;
Fontaine, f. et al.). Ceci est probablement le résultat d’usages fréquents dans l’élevage et le
traitement des porcelets d’agents antimicrobiens (Bertchinger, H. U. and Fairbrother, J. M.
1999; Hampson, D. J. 1994). Le pourcentage de souches résistantes est très élevé pour les
agents antimicrobiens suivants : tétracycline, sulfamides et spectinomycine.
Les infections extraintestinales causées par E. cou (ExPEC) sont communes chez les animaux
et chez les humains et peuvent être localisées dans plusieurs organes (Russo, T. A. and
Jolmson, J. R. 2000). Les infections typiques aux ExPEC incluent les infections urinaires
(UTI), les méningites, diverses infections intra-abdominales, des pneumonies, des infections
des muscles lisses, septicémies et ostéomyélites (Russo, T. A. and Jolmson, J. R. 2000). Les
ExPEC sont la cause la plus commune d’infections urinaires (UTI) chez les animaux de
compagnie et chez l’humain. La thérapie aux agents antimicrobiens est un outil important
dans le traitement de ces infections. Cependant, la résistance aux agents antimicrobiens
utilisés est étendue et concerne autant la médecine vétérinaire que la médecine humaine
(Aubry-Damon, H. and Courvalin, P. 1999; Monroe, S. and Polk, R. 2000; Teshager, T. et al.
2000).
Même si E. cou est associé à plusieurs types de maladies, un haut degré de spécificité existe
entre l’hôte et le type d’infection pour les différents sérogroupes. Certains sérogroupes
prédominent dans les infections chez les animaux, tandis que d’autres sont retrouvés
seulement dans les infections chez l’humain (Aarestrnp, F. M. and Wegener, H. C. 1999).
Toutefois, les ExPEC peuvent être isolés chez plusieurs différents hôtes. Des souches E. cou
isolées d’UTI canine et humaine semblent montrer beaucoup de similitudes génotypiques et
phylogéniques. Ces observations permettent de croire que les souches isolées du chien ne
feraient pas partie d’une population à part des souches isolées de l’humain (Johnson, J. R. et
aÏ. 2001; Feria, C. et al. 2001). Néanmoins, il semblerait, que ces similitudes ne sont pas
rencontrées lorsque la souche E. cou est isolée de lésions cutanées. Caya et al (Caya, F. et al.
1999) ont comparé des souches E. cou provenant de lésions sur des poulets et de différents
organes humains. Les résultats obtenus n’ont pas permis de démontrer qu’il existe entre ces
souches certaines similitudes qui penuetteraient de les mettre toutes dans un seul et même
groupe.
Les souches E. cou extraintestinales peuvent se distinguer des souches commensales E. cou et
des souches pathogènes intestinales selon leur lien phylogénétique (Johnson, J. R. and Russo,
T. A. 2002). Les souches commensales dérivent des groupes phylogénétiques A ou 31,
définis par l’électrophorèse des enzymes de base de la souche (Jolmson, J. R. and Russo, T.
A. 2002). De plus, on a démontré que les souches appartenant au groupe phylogénétique A ou
B possèdent moins de facteurs de virulence que les souches causant des infections (Herzer, P.
J. et aÏ. 1990). Les différentes souches pathogènes intestinales, qui donnent rarement des
infections extraintestinales, dérivent des groupes phylogénétiques A, 31 ou D ou encore
d’une lignée non regroupée. Celles-ci possèdent des facteurs de virulence caractéristiques liés
à des syndromes diarrhéiques. Par exemple, E. cou 0157: H7 ainsi que d’autres E. cou
16
entérohémonagiques (EHEC) produisent des shigatoxines, des adhésines intimines et des co
facteurs leur conférant une adhérence de type «attachement-effacement» caractéristique.
Contrairement aux souches commensales ou intestinales E. cou, les ExPEC dérivent du
groupe phylogénétique B2 et, plus rarement, du groupe D. Les facteurs de vimlence
caractéristiques des ExPEC incluent diverses adhésines, polysaccharides, toxines,
sidérophores, protéases, invasines et des protéines.
17
4.0 L’intégron
La résistance bactérienne non croisée à plus de trois agents antimicrobiens est courante dans
le monde de la médecine vétérinaire et humaine. On appelle ce phénomène la multirésistance.
Même si la sensibilité à un agent antimicrobien peut revenir après l’arrêt de l’utilisation de
celui-ci, les bactéries résistantes à cet agent antimicrobien peuvent persister longtemps après
la baisse des pressions sélectives. La résistance aux agents antimicrobiens peut persister car
celle-ci est unie à des liens génétiques créés par les changements occasionnés par les
pressions sélectives. L’habileté des bactéries à acquérir et à disséminer des gènes via des
éléments génétiques mobiles comme les plasmides et les transposons a été un aux agents
antimicrobiens durant les 50 dernières années. Vers la fin des années 80, un autre facteur
majeur dans le développement de la multirésistance a été découvert. Ce mécanisme implique
un élément d’ADN, appelé intégron, qui permet l’intégration de gènes de résistance à un site
spécifique par un événement de recombinaison.
Les intégrons sont impliqués dans la propagation, parmi les bactéries à Gram-négatif et
spécialement parmi les bactéries entériques, de plusieurs gènes de résistance aux agents
antimicrobiens. Les intégrons contierment généralement un gène codant pour une intégrase de
la famille de 2. (int]), un site d’intégration de cassettes (att]), où les gènes de résistance sous
forme de cassettes peuvent s’intégrer, et un promoteur responsable de l’expression des
cassettes insérées. L’intégrase de l’intégron recombine les cassettes en aval du site proximal
attl du promoteur. Une cassette contient une région codante (gène de résistance) et un élément
court de 59 paires de bases (pb) répété inversé contenant un site de recombinaison.
L’intégration de la cassette au site attl produit la formation d’un deuxième site d’insertion
(attC) en aval de la cassette. Jusqu’à maintenant, plus de 40 cassettes ont été décrites et quatre
classes d’intégrons sont connues (Recchia, G. D. and Hall, R. M. 1995). Les différentes
classes d’intégrons se distinguent par leurs gènes int. C’est l’intégron de classe 1 qui est le
plus souvent retrouvé parmi les souches cliniques d’entérobactéries.
Les intégrons de classe 1 contiennent à l’extrémité 3’ un gène codant pour une résistance à
l’ammonium quaternaire nommé qacEAl ainsi qu’un gène codant pour une résistance aux
18
sulfamides, suil (Figure 1). Les régions conservées 5’ et 3’ entourent une région variable
(RV) qui contient la ou les cassette(s). Les intégrons acquièrent ou échangent les gènes de la
RV en utilisant l’intégrase codée par le gène inti, qui est capable d’insérer et de faire la
recombinaison d’ADN mobile (cassette) à un site spécifique contenant une séquence
consensus (GTTRRRY). Cette intégrase est responsable de la liaison de gènes de résistance
permettant de former un large locus contenant de multiples gènes de résistance. L’intégron de
classe 1 a été détecté pour la première fois sur des dérivés complets ou tronqués du transposon
«Mu-like» Tn402 qui réside sur un plasmide ayant une spécificité d’hôte étendue ainsi que
parmi le transposon Tn21 ou «Tn21-like ». Tn21 est un grand transposon de 19,7-kb de
classe 2 qui transporte avec lui un opéron codant pour une résistance au mercure (mer), un
intégron de classe I (1n2) et un module de transposition (tnpA, tnpR et un site res). Les
différentes variantes du transposon Tn21 possèdent certaines cassettes ajoutées et/ou des
séquences d’insertion (IS) à la région 3’ insérées ou délétées. Le prototype de Tn2] porte le
gène de résistance à la spectinomycine et à la streptomycine, aadA], sur son intégron en plus
du gène merA qui confère la résistance au mercure. La large distribution du transposon Tn2]
parmi les isolats cliniques est possiblement attribuable aux plasmides conjugatifs sur lesquels
il réside. L’usage des sulfamides pendant six décennies et l’utilisation de la combinaison
triméthoprime-sulfamide ont entraîné une forte sélection en faveur des intégrons de classe 1
qui portent à leur extrémité 3’ le gène sulI.
19
à?P qacEAl sutI p orf5
I 5’.. I
____t
tut pGTTRRRY GTTRRRY
—s
Cassette insérée—J
Région 5 ‘-conservée Région 3 ‘-conservée
Figure 1. Structure générale de l’intégron de classe 1. Les flèches indiquent la direction de la
transcription. La direction et l’emplacement des différents promoteurs (Pi, F2 et F) sont
indiqués. La région GTTRRRY est la séquence d’intégration des cassettes. Les
oligonucléotides de 5’-CS et 3’-CS sont spécifiques à la région conservée 5’ et 3’
respectivement. Une cassette insérée dans la région variable est montrée avec ses 59-pb
conservées illustrées, ici, par un bloc noir. (Adaptée et tirée de (Levesque, C. et al. 1995))
20
5.0 La génétique de la résistance
Certaines espèces bactériennes possèdent une résistance intrinsèque à des agents
antimicrobiens donnés. Cette résistance est observée dès l’utilisation de l’antibiotique. Par
exemple, les mycoplasmes sont résistants à la pénicilline. Cependant, ce n’est pas ce type de
résistance qui cause problème dans le traitement des infections bactériennes. Les résistances
qui sont acquises par la bactérie représentent la problématique principale. La résistance peut
provenir d’une ou des mutations ponctuelles dans certains gènes chromosomiques, mais ces
événements surviennent très rarement (1 sur 10g) chez les bactéries. Puisque chez E. cou, la
résistance acquise à la faveur de nouveaux gènes est la plus courante pour les agents
antimicrobiens considérés dans cette étude, la résistance aux agents antimicrobiens due à des
mutations chromosomiques sera peu discutée dans ce mémoire. Ce qui sera plus largement
discuté ici aura trait aux résistances aux agents antimicrobiens qui résulte de l’acquisition de
gènes de résistance par la bactérie.
Plusieurs mécanismes de résistance existent. Certains gènes vont permettre la production
d’enzymes qui vont inactiver l’agent antimicrobien. D’autres vont permettre la production de
protéines transmembranaires jouant un rôle de pompe cellulaire. Ces pompes vont excréter
spécifiquement les différents agents antimicrobiens hors de la cellule. La bactérie peut aussi
produire une nouvelle cible pour l’agent antimicrobien ou, simplement, modifier la cible
originale.
La bactérie acquière une ou des résistances à différents agents antimicrobiens par l’acquisition
de nouveaux gènes via les éléments génétiques mobiles (transposons, intégrons, plasmides,
etc.). En fait, plusieurs gènes peuvent être transmis par le même élément génétique.
La résistance croisée est un phénomène commun entre les membres d’une même famille. Par
exemple, une bactérie résistante à l’oxytétracycline l’est automatiquement à la
chlortétracycline et la tétracycline. La résistance croisée se différencie de la résistance
multiple par le fait qu’elle se manifeste chez une souche vis-à-vis plusieurs antibiotiques
appartenant à des familles différentes. Alors que la résistance croisée implique un mécanisme
21
de résistance commun, la résistance multiple est généralement associée à plus d’un
mécanisme. Par exemple, la résistance à lampicilline, la tétracycline et le chloramphénicol
chez une même souche relève de la résistance multiple.
Plusieurs gènes de résistance permettant à la bactérie qui les acquière de développer des
résistances à plusieurs agents antimicrobiens ont été décrits dans la littérature. Puisque le sujet
de cette maîtrise porte sur la résistance aux agents antirnicrobiens retrouvée chez E. coït, seuls
les gènes de résistance acquis, et donc mobiles, qui sont les plus souvent retrouvés chez cette
bactérie, seront décrits (Tableau II).
22
Tableau II. Mécanismes d’action des déterminants génétiques de résistancea
Localisation Nomenclature NumérosEnzymegénétigue’
Gènealternative
Mécanisme d’action Profil de résistanced’accession
TEM-1 p tem-1- AMP, AMK, PIP, TIC AF309524
SHV-1 p shv-1 pit-2 AMP, AMK, PIP, TIC AF148850
AMP, AMK, AMX, PIP,OXA-1 p oxa-1 - M235349TIC
Hydrolyse le noyau beta-lactamAMP, AMK, AMX, PIP,OXA-7 p oxa-7
- des beta-lactamines X75562TIC
PSE-4 chr, p, tn pse-4- AMP, CAR, PIP, TIC J05162
AMP, AMX, ATM, CTX,X52556CTX-M-3 P ctx-m-3
- CR0, PIP, TIC, TIM
Modifie la fonction hydroxylANT(2”)-I p ant(2’9-!a aadB des aminoglycosides par une GEN, KAN, T0B X04555
adénylylation
aacC2, aacC3,MC(3)-II p aac(3)-!a Modifie la fonction amino des GEN, T0B X54723aacC5, aac(3)-Vaaminoglycosides par une
AAC(3)-IV p aac(3)-!Va - acélylation APR, GEN, TOB X01385
APH(3’)-I tn aph(3’)-ia aphA-1 Modifie la fonction hydroxyl < NEO M18329
des aminoglycosides par uneAPH(3’)-ll tn aph(3’)-i!a aphA-2 phosphorylation KAN, NEO Vomis
Tet(A) p tef (A) - TET X00006
Tet(B) p tet(B)- MIN, TET J01830
Tet(C) p tet(C)- Protéine de membrane TET J01749
exportant la tétracyclineTet(D) p tet(D)- TET X65876
Tet(E) p tet(E) - TET L06940
Tet(Y) p tet(Y)- TET AF070999
CAT1 p, tn cati- CHL M62822
Catalyse l’scétylation du 0H-3’CHL X53796CAT1 P tn cati!
- du chloramphénicol
CAT11 p, tn cati!! - CHL X07848
Protéine de membraneFI0R chr, p foR flo exportant le chloramphénicol et CHL, FLO AF252855
le florfénicol
Dhfr I chr, p, tn dhfri dfrl TMP X00926
Dhfr V chr, p, tn dhfrV dfr5 TMP X12868Variant de l’enzyme cible
Dhfr VII chr, p, tn dhfrVi! dfr7 dihydrofolate insensible à la TMP X58425
triméthoprimeDhfr IX chr, p, tn dhfr!X dfr9 TMP X57730
Dhfr XIII chr, p, tn dhfrXIiI dfrl3 TMP Z50802
SuIl p, tn su!! su!; Variant de l’enzyme cible SUL X12s69
dihydropteroate insensible auxSulIl p su!!! su!2 sulfamides SUL M35657
a Ces informations ont été tirées des références suivantes: (Ambler, R. P. et al. 1991; Bush,
K. et al. 1995; Du Bois, S. K. et aÏ. 1995; Yang, Y. et al. 1999; Kotra, L. P. et al. 2000;
Mingeot-Leclercq, M. P. et aÏ. 1999; Shaw, K. J. et al. 1993; Wright, G. D. et al. 1998;
23
Chopra, I. and Howe, T. G. 1978; Chopra, I. and Roberts, M. 2001; Huovinen, P. et aÏ. 1995;
Skold, 0. 2001).b chr, chromosomique; p, plasmidique; tn, transposon.
AMP, ampicilline; AMK, amoxicilline; AMX, amoxicilline-acide clavulanique; ATM,
aztreonarn; APR, apramycine; CAR, carbenicilline; CR0, ceftriaxone; CTX, cefotaxime;
CHL, chloramphénicol; FFL, florfénicol; GEN, gentamicine; KAN, kanamycine; MiN,
minocycline; NEO, néomycine; PIP, piperacilline; SUL, sulfamides; TET, tétracycline; TIC,
ticarciHine; TIM, ticarcilline-acide clavulanique; TMP, friméthoprime; TOB, tobramycine.
24
6.0 Les agents antimicrobiens
6.1 Les 3-lactamines
Les 13-lactamines restent toujours la famille d’agent antimicrobiens la plus utilisée, la plus
diversifiée et la moins toxique. Elle reste en pleine évolution grâce aux nouveaux produits qui
paraissent continuellement.
Depuis la mise en marché de la pénicilline G en 1942, on a assisté à l’apparition de la
première pénicilline oral, la pénicilline V. Par la suite, à partir de l’obtention de l’acide 6-
aminopénicillanique, le noyau commun de toutes les pénicillines, et de l’acide 7-
aminocéphalosporanique, le noyau commun de toutes les céphalosporines, il y a eu
l’apparition successive des f3-lactamines semi-synthétiques et des céphalosporines de
première, deuxième, troisième et quatrième génération. Il s’en suit de la synthèse des
monobactames, des pénèmes et carbapénèmes et finalement, des inhibiteurs irréversibles de
Ç3-lactamascs (Neuman, M. 1990). Le développement des aminopénicillines, qui indu
l’ampicilline, a permis de traiter les infections causées par la bactérie E. cou, laquelle est
naturellement résistante à la pénicilline.
Les f3-lactamines peuvent être classées selon plusieurs critères. Le plus employé fait appel à
leur mode d’obtention, leur composition chimique et leur indication principale. Les 3-
lactamines vont fixer préférentiellement les cibles (récepteurs) spécifiques, que sont les
«penicillin binding proteins » (P.B.P.). Ces P.B.P., mises en évidence par Spratt en 1975 sur
un mutant de E. coÏi non producteur de 13-lactamases, sont situées sur la membrane interne de
la paroi bactérienne (Neuman, M. 1990). Celle fixation entraîne l’arrêt de la synthèse du
peptidoglycan et donc de la croissance bactérienne, ce qui correspond à l’effet
bactériostatique de la f3-lactamine. L’effet bactéricide des -lactamines résulte de la libération
d’autolysines par la bactérie, ce qui provoque l’éclatement cellulaire de la bactérie. Le mode
d’action des 3-lactamines explique leur inactivité sur les bactéries sans paroi ou intra
cellulaire Mycoplasma, ChÏarnydia, Rickettsia.
25
Mécanismes d’action des 13-lactamases
La production de 3-lactamases qui hydrolysent le noyau f3-lactam et inactivent l’agent
antimicrobien engendre la résistance aux 3-lactamines est le mécanisme le plus fréquent
(Livermore, D. M. 1995). Plusieurs différentes f3-lactamases ont été décrites (Bush, K. et al.
1995; Livermore, D. M. 1995). Plus de 200 13-lactamases ont été classifiées en trois groupes
principaux et huit sous-groupes, selon leur fonction et leurs caractéristiques structurales
(Bush, K. et al. 1995). Leur mode d’action varie légèrement d’une enzyme à une autre mais,
fondamentalement, il repose sur le fait que l’enzyme se lie de façon non covalente aux f3-
lactamines. Par la suite, le noyau f3-lactam est attaqué par des molécules hydroxyle libres
générant un lien covalent. Finalement, l’enzyme active est libérée par une hydrolyse au niveau
du lien covalent et ceci inactive la f3-lactamine (Livermore, D. M. 1995). Chez les bactéries à
Gram-négatif, ces enzymes sont produites et restent emprisonnés dans t’espace périplasmique
de la bactérie, contrairement aux bactéries à Gram-positif qui excrètent plutôt les f3-
lactamases (Livermore, D. M. 1995).
Les 3-lactamases sont divisées en trois groupes selon leurs caractéristiques fonctionnelles. Le
premier groupe contient les céphalosporinases non inhibées par l’acide clavulanique; le
deuxième groupe comprend les pénicillinases, les céphalosporinases et la Ç3-lactamases à large
spectre qui sont généralement affectées par les inhibiteurs de 3-lactamases; et le troisième
groupe regroupe les metallo-f3-lactamases qui hydrolysent les pénicillines, les céphalosporines
et les carbapénèmes. Une autre classification a été proposée; celle-ci fait appel à la structure
des enzymes. La classe A contient toutes les f3-lactamase de type pénicillinases; la classe B,
les metalo-f3-lactarnases; la classe C regroupe les céphalosporinases; et finalement la classe D
les enzymes hydrolysant les oxacillines (Bush, K. et al. 1995). Parmi ces classes, il existe des
3-lactamases qu’on appelle des «Extended-spectrum 3-lactamase» ou ESBL et qui ont
comme caractéristique de rendre la bactérie résistante à plusieurs céphalosporines de
troisième génération.
E. cou représente le seul cas, avec Shigella spp., où une enzyme de classe C chromosomique,
non inductible, de type AmpC, peut être impliquée dans la résistance aux 3-lactamines
26
(Jacoby, G. A. and Caneras, I. 1990). La bactérie peut produire en grande quantité des
enzymes de type AmpC et devenir résistante. Toutefois, cette enzyme n’est pas le seule mode
de résistance retrouvée chez E. cou. Généralement, la résistance provient de l’acquisition de
gènes de résistance via l’acquisition de plasmides.
Les enzymes les plus fréquemment isolées sont de type TEM, SHV, OXA, PSE et plus
rarement, de type CTX-M. Les gènes codant ces enzymes sont très nombreux. Les variantes
des gènes tei;z ainsi que shv, pse et ctx-m se différencient par un ou deux changements dans la
séquence nucléotidique. Ces mutations transforment la structure tri-dimensionnelle des
enzymes et changent par le fait même leur spécificité enzymatique (Du Bois, S. K. et aï.
1995). Présentement, 92 variantes de tem, 26 de shv, 40 de type oxa, 4 depse et 15 de ctx-m
ont été décrites (Perilli, M. et al. 2002; Chang, F. Y. et al. 2001; Lopez-Otsoa, F. et aÏ. 2002;
Savoie, A. et aÏ. 2000; Poire!, L. et al. 2002).
Les différentes variantes de ces déterminants ont des spectres d’action différents selon les
mutations ponctuelles. Il en résulte donc un spectre d’action élargi et I’inactivation de
certaines céphalosporines. Toutefois, la plupart reste tout de même inactive contre les
inhibiteurs de 3-lactamases (Livermore, D. M. 1995). La sensibilité à ces inhibiteurs peut être
modifiée par différents mécanismes. Le mécanisme le plus fréquent est l’hyperproduction ou
la synthèse d’une 3-lactamase résistante aux inhibiteurs de F3-lactamase de type TEM. Une
autre possibilité est l’hyperproduction d’une f3-lactamase chromosomique ou plasmidique de
type AmpC, par une amplification de gènes ou l’introduction de mutations au niveau du
promoteur. La production d’une r3-Iactamase de type OXA peut aussi conduire à une
résistance aux inhibiteurs (Livermore, D. M. 1995).
Il est à noter que les gènes tein sont les gènes les plus fréquemment retrouvés chez E. cou. Les
gènes shv, qui semblent dériver de Kiebsiella spp., montrent environ 65% d’identité avec les
gènes tein (Bush, K. et al. 1995). Les régions non identiques dans leur séquence permettent de
les différentier lors d’études biomoléculaires.
27
Distribution et mobilité de la résistance
Depuis les 50 dernières années, la résistance aux f3-lactamines, transmise grâce à des
plasmides, est devenue courante chez les entérobactéries et autres bactéries à Gram-négatif.
Les enzymes de type TEM, $11V, OXA, PSE et CTX sont produites par des gènes portés par
des plasmides. L’expression de la Ç3-lactamase TEM-l est le mécanisme de résistance le plus
commun chez les bactéries à Gram-négatif (Bush, K. et al. 1995; Chaibi, E. B. et al. 1999).
Le gène codant pour cette enzyme est localisé sur le transposon Tn3, quoique celui-ci et ses
dérivés puissent être retrouvés sur Tn2 (Rasmussen, B. A. et al. 1993). Le transposon où
réside le gène tem-1 est certainement plus mobile que l’ADN qui contient le gène shv-1, ce
qui expliquerait que les gènes tem-J sont plus souvent retrouvés que shv-]. Toutefois, ce
phénomène n’a pas été clairement démontré.
Utilisation des 3-Lactamines
Les 13-lactamines sont utilisées largement afin de traiter des infections chez l’animal et chez
l’humain. L’augmentation de l’utilisation ainsi que l’usage répété des Ç3-lactamines ont
conduit vers une résistance fréquente à ces agents antimicrobiens. Deux stratégies ont alors
été employées afin de contrer cette résistance. Premièrement, de nouvelles f3-lactamines qui
étaient moins touchées par les 3-lactamases ont été développées. Deuxièmement, l’utilisation
de la combinaison d’inhibiteurs de 13-lactamases, comme l’acide clavulanique, le sulbactame
ou le tazobactame, de concert avec une pénicilline a aussi été privilégiée afin de rendre
efficace les traitements aux 3-lactamines. Toutefois, vers 1990, l’efficacité de ces agents
inhibiteurs a chuté à cause de l’émergence de variantes de l’enzyme de type TEM qui rendent
les bactéries résistantes à plusieurs de ces agents (Thomson, K. S. and Smith Moland, E.
2000).
6.2 Les aminoglycosides
Malgré tous les efforts pour remplacer les aminoglycosides, en raison de leurs inconvénients
réels (marge thérapeutique étroite, à cause de leur toxicité pour les reins et les oreilles et
conséquemment, la nécessité fréquente de monitoring des taux sériques, en plus de la
nécessité d’administration parentérale à cause de leur faible absorption au niveau gastro
intestinal), ceux-ci restent toujours indispensables et continuent leur carrière surtout en milieu
28
hospitalier dans le traitement des infections nosocomiales sévères, en association avec une 3-lactamine.
Les aminoglycosides ont une action bactéricide, dose-dépendante, sur les bactéries par
l’altération de la synthèse protéique de celles-ci par action directe sur leur ribosome. L’agent
se fixe sur la protéine Si2 et sur d’autres sites des sous-unités 30S des ribosomes avec, comme
conséquence, la formation de peptides aberrants non fonctionnels par transcription incorrecte
de l’information par l’ARN messager (erreur de lecture) et l’inhibition de la transiocation
(Neuman, M. 1990). Le passage transmembranaire des aminoglycosides dans la bactérie est
un phénomène actif qui requiert de l’oxygène et, de ce fait, il ne peut intervenir en
anaérobiose. Ce qui explique la résistance naturelle des bactéries anaérobies strictes aux
aminoglycosides.
Mécanisme d’action de la résistance aux aminoglycosides
Chez les bactéries, la résistance aux arninoglycosides fait souvent suite à l’inactivation
enzymatique par des acétyltransférases (AAC), adénylyltransférases (ou
nucléotidyltransférases) tANT) et des phosphotransférases (APH) (Shaw, K. J. et aL 1993).
Chacun de ces enzymes affecte différents membres de la famille des aminoglycosides. En
plus de la résistance enzymatique, des altérations du ribosome ainsi qu’une perte de
perméabilité chez la bactérie peuvent amener une résistance aux aminoglycosides. Toutefois,
à l’exception de la résistance à la streptomycine ou à la spectinomycine, il semblerait que la
résistance soit due plus fréquemment à une acquisition de gènes de résistance via des
plasmides que par des mutations au niveau du chromosome (Shaw, K. J. et aÏ. 1993).
La nomenclature des enzymes impliquées dans la résistance aux aminoglycosides est établie
comme suit ($haw, K. J. et al. 1993)
1- Le type de modification enzymatique est déterminé (AAC (acétyltransférases), ANT
(adénylyltransférases) ou APH (phosphotransférases))
2- Le site de la modification est nommé ((1), (3), (6), (9), (2’), (3’), (4’), (6’), (2”), ou
(3”))
29
3- Pour le profil de résistance, I, II, III, IV, etc. est utilisé suivit de a, b, c, d, etc. pour
l’unique désignation enzymatique
Ainsi, AAC(6’)-Ia et AAC(6’)-Ib sont deux enzymes distintes, conférant un profil de
résistance identique. Les gènes codant ces enzymes portent la même nomenclature à
l’exception que les lettres sont toutes en minuscules et le nom est en italique. Pour suivre
l’exemple précédant, les gènes codant pour ces deux enzymes se nomment aac(6 9-la et
aac(6 9-lb. D’autres nomenclatures sont aussi employées, mais celle-ci est la plus
fréquemment utilisée.
En raison de la grande diversité des gènes de résistance aux aminoglycosides, on compte 33
mécanismes de résistance et plus de 70 gènes décrits (Shaw, K. J. et al. 1993). Seuls les gènes
retrouvés chez E. cou seront discutés dans le présent mémoire. Parmi les gènes de résistance
aux aminoglycosides codant pour un acétyltranférase, le gène aac(3)-Ia qui procure une
résistance à la gentamicine est retrouvé chez 10 à 17% des bactéries à Gram-négatif. Le gène
aac(3)-IIa qui procure entre autres une résistance à la gentamicine est retrouvé chez 60,3%
des bactéries à Gram-négatif. f inalement, le gène aac(3)-IVa, conférant une résistance à la
gentamicine ainsi qu’à l’apramycine, n’est retrouvé que rarement (3,5%) chez les
Enterobacteriaceae ($haw, K. J. et al. 1993). Le gène ant(2’9-Ia est l’un des gènes les plus
communs chez les bactéries à Gram-négatif (15%) codant pour une enzyme ayant une activité
de 2”-O-adénylyltransférase. Celui-ci code pour une résistance à la gentamicine et à la
kanamycine. Parmi les gènes codant pour une enzyme ayant une activité de phosphorylation,
aph(3 9-la, retrouvé sur le transposon Tn903, est le plus fréquemment retrouvé (46%) parmi
les bactéries à Gram-négatif avec le gène aph(3 9-lia retrouvé parfois chez E. cou. Tous les
deux confèrent une résistance à la néomycine ainsi qu’à la kanamycine (Shaw, K. J. et al.
1993).
Distribution et mobilité des gènes de résistance
Les gènes de résistance qui codent pour une enzyme qui modifie les aminoglycosides par une
acétylation, adénylation ou phosphorylation, constituent des gènes fréquemment rencontrés
sous forme de cassettes dans la RV des intégrons. Les gènes de résistance codant pour une
enzyme de type ANT(3”) sont retrouvés sous forme de cassettes dans la nature. Le gène
30
ant(3 ‘)-Ia (aadA]) qui a été décrit pour la première fois par Sundstrôm et aï (Sundstrom, L.
et al. 198$), est souvent retrouvé sur le transposon Tn21.
Il existe plus de 20 gènes de résistance à la gentamicine qui sont décrits jusqu’à maintenant.
Le gène aac(3)-IIa (aussi nommé aacC2), ant(2’9-Ia (aadB) et aac(3-Ia (aacCl) sont
souvent retrouvés parmi les souches à Gram-négatif isolées d’humains (Shaw, K. J. et aï.
1993; Miller, G. H. et al. 1997). Le gène aac(3)-IV qui donne une résistance à la gentamicine,
tobramycine et Ï’apramycine a été retrouvé chez des isolats provenant d’animaux (Wray, C. et
al. 1993; Hunter, J. E. et al. 1994). II semblerait qu’aucun autre gène de résistance à la
gentamicine, sauf ces quatre gènes, n’a été isolé chez l’animal ($andvang, D. and Aarestnip,
F. M. 2000). Tous les gènes de résistance à la gentamicine donnent une résistance croisée à
d’autres aminoglycosides.
Utilisations des aminoglycosides
Les aminoglycosides sont utilisés le plus souvent chez les humains en milieu hospitalier. Elles
sont destinées aux traitements d’infections nosocomiales et presque toujours en association
avec un deuxième agent. En pratique extra-hospitalière, les indications sont rares, à
l’exception de la spectinomycine dans le traitement des gonococcies résistantes aux 13-lactamines. Les aminoglycosides servent à traiter les septicémies, les endocardites et autres
infections graves (Neuman, M. 1990).
À cause de la toxicité des aminoglycosides et de leur capacité à rester longtemps dans les
tissus, l’utilisation de ces agents chez les animaux par voie parentérale a été largement
réduite. Leur usage est maintenant réservé pour les septicémie ou les infections systémiques.
Le premier choix est alors la gentamicine, suivi de l’amikacine et de la tobramycine (Prescott,
J. F. et al. 2000). Toutefois, les aminoglycosides tels que la néomycine et l’apramycine
restent encore utilisés par voie orale afin de contrôler les diarrhées causées par les
entérobactéries (Prescott, J. F. et aÏ. 2000).
31
6.3 Les tétracyclines
Les tétracyclines, qui ont été découvertes dans les années 40, appartiennent à une famille
d’agents antimicrobiens qui inhibent la synthèse protéique par l’empêchement de
l’attachement de l’aminoacyl de l’ARN de transfert au ribosome. Les tétracyclines sont des
agents antimicrobiens à large spectre. Leur activité touche plusieurs bactéries à Gram-positif
et à Gram-négatif, mais aussi les bactéries atypiques comme les ChÏamydia, les Mycoplasnza,
et certains protozoaires (Chopra, I. and Roberts, M. 2001). Les propriétés avantageuses de ces
agents antimicrobiens et l’absence d’effets secondaires majeurs ont mené à une utilisation
importante des tétracyclines dans le traitement des infections chez l’animal et chez l’humain.
De plus, en Amérique du nord, ceux-ci sont utilisés à des fins autres que thérapeutiques chez
les animaux de consommation comme facteurs de croissance. Même si les tétracyclines
présentent beaucoup d’intérêts pour la médecine vétérinaire et humaine, l’émergence de
microorganismes résistants limite leur efficacité. Les efforts qui ont été faits pour améliorer la
compréhension de la résistance aux tétracyclines ont permis la découverte d’une nouvelle
génération de tétracyclines, les glycylcyclines (Chopra, I. and Roberts, M. 2001).
La chlortétracycline et l’oxytétracycline ont été découvertes à la fin des années 40. Ces
molécules sont produites par Streptornyces aureofaciens et S. rimosus, respectivement.
D’autres, moins utilisées, proviennent aussi des Streptornyces ou encore d’une production
semi-synthétique. La glycylcycline, la dernière des tétracyclines à être introduite, est d’origine
semi-synthétique (Chopra, I. and Roberts, M. 2001).
Mécanisme d’action des gènes de résistance à la tétracycline
La résistance aux tétracyclines est un phénomène relativement répendu et découle de son
utilisation intensive dans l’agriculture et en médecine humaine. La résistance aux
tétracyclines émerge chez les bactéries commensales et pathogènes suite à l’acquisition des
gènes tet. Vingt-neuf gènes de résistance à la tétracycline tet et trois gènes de résistance à
l’oxytétracycline (ofr) ont été caractérisés. Certains de ces gènes codent pour des pompes à
efflux et d’autres codent pour une protection ribosomale. Les pompes sont les plus
documentées et étudiées des protéines Tet. Tous les gènes tet codant pour les pompes, codent
pour des protéines associées à la membrane qui exportent les molécules de tétracycline de
32
l’intérieur de la cellule à l’extérieur (Chopra, I. and Roberts, M. 2001). Ces gènes sont
retrouvés chez les bactéries à Gram-positif et à Gram-négatif. La majorité de ces protéines Tet
confèrent une résistance à la tétracycline, mais non à la minocycline et à la glycylcyclines.
Seul tet(B) confère la résistance à la tétracycline et à la minocycline. Les gènes tet codant
pour une pompe retrouvée chez les bactéries à Gram-négatif sont normalement associés avec
des plasmides conjugatifs. Les gènes tet retrouvés chez E. cou sont tet(A), tet(B), tet(C),
tet(D), tet(E), tet(I) et tet(Y) (Chopra, I. and Roberts, M. 2001). Tous font partie des gènes
codant pour des pompes. Il n’y a donc aucun gène tet retrouvé chez E. cou qui puisse conférer
une protection ribosomale. Les gènes tet codant pour une pompe retrouvés chez les bactéries à
Gram-négatif sont régulés par la tétracycline. En l’absence de tétracycline, la transcription de
ces gènes est bloquée (Chopra, I. and Roberts, M. 2001).
De plus, d’autres mécanismes de résistance non plasmidique ont été décrits. Certaines
résistances sont dues à des mutations au niveau du chromosome (mar locus) de la bactérie ou
encore au niveau de la région 16S de l’ARN ribosomal (Linde, H. J. et al. 2000; Cohen, S. P.
et al. 1993; Ross, J. I. et al. 1998).
Distribution et mobilité des gènes tet
On retrouve les gènes tet chez une variété de bactéries isolées d’animaux et d’humains. La
majorité des gènes tet sont associés à des éléments génétiques mobiles. Les gènes tet
retrouvés chez les bactéries à Gram-négatif et codant pour des pompes à efflux sont retrouvés
sur divers groupes de plasmides conjugatifs pouvant avoir des incompatibilités de groupes
(Jones, C. S. et al. 1992). Deux plasmides s’excluant mutuellement, c’est-à-dire ne pouvant
coexister dans la même bactérie, appartiennent au même groupe d’incompatibilité. Jones et al
ont trouvé une corrélation entre l’incompatibilité de certains plasmides et la présence de
certains gènes tet sur ceux-ci. Ils suggèrent que certains de ces gènes sont liés génétiquement
par leur incompatibilité. Le gène tet(E) diffère des autres gènes codant pour une pompe à
effiux, car il est associé à un gros plasmide qui n’est ni mobile ni conjugatif (Allard, J. D. and
Bertrand, K. P. 1993). Ceci peut expliquer sa distribution et sa prédominance limitée.
33
Utilisation des tétracyclines
Les tétracyclines ont été largement employées en médecine humaine. Toutefois, avec le
développement de résistance dans un pourcentage élevé d’entérobactéries et de cocci à Gram-
positif et anaérobes (Streptococcus, Staphytococcus et Enterococcus), les indications actuelles
de tétracyclines se limitent aux traitements des infections à bactéries intracellulaires
(BrucelÏa, Borrelia, Chlarnydia, MycopÏasrna, Rickettsia et Yersinia) (Neuman, M. 1990).
Chez l’animal, les tétracyclines sont employées dans le traitement d’infections chez les porcs,
la volaille, les vaches et les moutons. Les tétracyclines représentent la classe d’agents
antimicrobiens la plus utilisée chez les animaux; ils représentent environ 50% (en quantité)
des agents antimicrobiens utilisés chez l’animal {Schwarz, 2001 #234]. Elles peuvent être
aussi proscrits pour le traitement de certaines d’infections chez les animaux de compagnie.
Les tétracyclines sont aussi utilisées à des fins autres que thérapeutiques chez les animaux de
consommation, afin d’augmenter leur taux de croissance et la conversion alimentaire. Les
propriétés subthérapeutiques des tétracyclines ont été découvertes en 1949. Ces effets
subthérapeutiques ont été observée chez la volaille, puis chez le porc et le bovin. Cependant,
ce n’est qu’en 1951 et 1953 que la chlortétracycline et l’oxytétracycline, respectivement, ont
été approuvées par la « Food and Drug Administration » (fDA) aux États-Unis comme additif
alimentaire chez les animaux. Ensuite, ces utilisations ont été mises en question par la
Grande-Bretagne qui suggérait que l’utilisation subthérapeutique des tétracyclines augmente
la résistance à ces agents chez l’animal et l’humain. En Europe, cette hypothèse a été prise au
sérieux. Il en est résulté le banissement de l’utilisation des tétracyclines comme facteur de
croissance chez les animaux, en 1970.
6.4 Les phénicols
Le chloramphénicol est un agent antimicrobien de large spectre qui était utilisé de façon
intensive avant son bannissement à cause de ses propriétés toxiques. En effet, l’humain
exposé à cet agent pouvait développer une aplasie médullaire irréversible (Neuman, M. 1990).
C’est donc par mesure de précaution que son usage chez les animaux de ferme a été banni au
Canada en 1980. Présentement, seule la forme fluorée, sous le nom de florfénicol, peut être
employée chez les animaux.
34
Le florfénicol est un composé fluoré analogue au chlorampliénicol et au thiamphénicol,
approuvé par la FDA en 1996 pour le traitement des infections causées par les pathogènes des
voies respiratoires bovines comme FasteureÏÏa spp. (White, D. G. et al. 2000). Cependant,
leur utilisation contre les infections entériques bovines à E. cou n’est pas encore approuvé aux
États-Unis. Son utilisation chez l’humain n’est pas acceptée. Le florfénicol possède un spectre
d’action similaire au chloramphénicol. Toutefois, il est actif à des concentrations moindres
que ce dernier, et ce, pour une vaste sélection de microorganismes, même ceux démontrant
une résistance au chloramphénicol (Syriopoulou, V. P. et al. 1981). Le florfénicol est
considéré comme un agent bactériostatique (White, D. G. et aÏ. 2000). Il agit directement sur
la synthèse protéique de la bactérie par une liaison de la molécule à la sous-unité 50$ du
ribosome.
Mécanisme d’action de la résistance
La résistance au chloramphénicol peut provenir d’une réaction enzymatique via une
inactivation chimique de l’agent ou à des pompes à efflux. Le chloramphénicol
acétyltransférase (CAT) catalyse l’acétylation du OH-3’ du chloramphénicol et est
responsable en majorité de la résistance à cet agent (Murray, I. A. and Shaw, W. V. 1997). Il
existe aussi une résistance non enzymatique au chloramphénicol provoquée par l’acquisition
du gène crnÏA (Bissonnette, L. et al. 1991). De plus, récemment, un nouveau gène,floR, a été
identifié sur un plasmide isolé d’une souche E. cou. Ce gène confère une résistance croisée au
chlorarnphénicol et au florfénicol due à l’action d’une pompe (Cloeckaert, A. et al. 2000).
Contrairement au gène foR, les gènes cat et cmM ne peuvent donner une résistance au
flofénicol.
L’acquisition de gènes de type cat est la cause majeure des résistances observées au
chloramphénicol. Ces gènes codant pour une chloramphénicol acétyltransférase sont dispersés
parmi plusieurs espèces bactériennes à Gram-positif et à Gram-négatif (Murray, I. A. and
Shaw, W. V. 1997). Trois variantes de ces gènes ont été très bien décrites dans la littérature
[Murray, 1997 #170]. Il s’agit de catI, catil et catili.
35
Distribution et mobilité de la résistance
Les gènes cat sont retrouvés sur plusieurs plasmides de type «f-like» liés ou associés à la
présence de transposons. Par exemple, le gène catI est retrouvé sur le transposon Tn9
(Munay, I. A. and Shaw, W. V. 1997).
Le gène foR est porté par des plasmides et/ou des transposons (White, D. G. et aÏ. 2000).
Cependant, la séquence de l’ADN entourant le gène JIoR retrouvé chez E. cou montre des
similitudes avec la séquence du plasmide RSF 1010 séquencé chez Fasteurella piscicida. Ceci
suggère que le gène de résistance au florfénicol et au chloramphénicol retrouvé chez E. cou
vienne d’un transfert entre deux espèces d’un plasmide ayant une spécificité d’hôte étendue
(White, D. G. et aÏ. 2000). De plus, une étude récente a démontré l’existence de ce gène sur le
chromosome de certaines souches E. cou (Doublet, B. et al. 2002). Ceci peut être expliqué par
le fait que différents sites d’intégration aux transposons existent sur le chromosome de la
bactérie.
Utilisation du chloramphénicol
Du fait de sa toxicité, le chloramphénicol n’est employé chez l’humain qu’en cas de sévères
infections, où ses qualités pharmacocinétiques de diffusion tissulaire, de concentration dans la
lymphe et de pénétration intracellulaire, le rendent encore indispensable. Le traitement des
salmonelloses systémiques, des méningites à Haernophilus influenzae et des mélioïdoses peut
justifier l’usage du chloramphénicol (Neuman, M. 1990).
L’usage du chloramphénicol en médecine vétérinaire est limité. C’est à cause de sa toxicité
chez l’humain que l’usage du chloramphénicol chez les animaux destinés à la consommation
humaine est maintenant banni, mais reste toujours accepté pour les animaux de compagnie et
pour le traitement d’infections systémiques à SalmoneÏÏa, d’infections profondes de l’oeil et
d’infections causées par des bactéries anaérobes (Prescott, J. f. et al. 2000).
6.5 La triméthoprime
La triméthoprime est un agent antimicrobien synthétique appartenant au groupe des composés
diaminopynmidine (Skold, 0. 2001). Cet agent a été utilisé pour la première fois en 1962 en
36
Angleten-e. La triméthoprime interfère de façon spécifique avec l’enzyme bactérienne, la
dihydrofolate réductase (DHfR). La triméthoprime a une structure analogue à la
dihydrofolate, ce qui lui permet d’être en compétition avec celle-ci pour la DHFR. Ceci
empêche la réduction de la dihydrofolate en tétrahydrofolate (Huovinen, P. et aÏ. 1995) qui est
nécessaire pour la synthèse de la thymine (Skold, 0. 2001). Il est à noter que l’enzyme DHFR
humaine est naturellement résistante à la triméthoprime. Ceci a été démontré par des études en
cristallographie aux rayons-X. Ces études ont montré que la triméthoprime s’intégrait bien
dans le site de liaison aux nucléotides de la DHfR bactérienne, mais aucunement à la DHfR
des mammifères (Matthews, D. A. et aÏ. 1985).
Puisque la triméthoprime est un agent antimicrobien synthétique, une résistance bactérienne à
cet agent n’avait pas été considérée possible. Par contre, 16 gènes exprimant une DHFR
résistante à la triméthoprime sont maintenant décrits chez différentes espèces bactériennes
(Skold, 0. 2001).
Mécanismes d’action de la résistance
La résistance à la triméthoprime chez les entérobactéries est devenue problématique vers les
années 1983 et 1984. Avant cela, moins de 4% des isolats d’entérobactéries étaient résistants.
Le mécanisme le plus commun de résistance à la triméthoprime chez les entérobactéries est la
production d’un plasmide qui produit une DHfR additionnelle et qui, contrairement à
l’enzyme chromosomique, est beaucoup moins sensible à la triméthoprime. Seize enzymes
procurant une résistance à la triméthoprime ont été identifiées et caractérisées. Celles-ci ont
été regroupées en deux groupes selon leur homologie de séquences. Le plus grand de ces deux
groupes contient les gènes que l’on retrouve le plus fréquemment et les enzymes de «type I
like », qui inclut dÏfrL dI!frlb, dhfrfrÇ dhfrVl et dIfr VII (Huovinen, P. et al. 1995). Une des
caractéristiques de ces enzymes est qu’elles portent un cadre de lecture ouvert (ORF) de 157
acides aminés (Huovinen, P. et al. 1995).
En plus de la présence de gènes de résistance pouvant être acquis par la bactérie, il existe des
résistances dues à des mutations dans le gène codant potir la DHfR chromosomique.
37
Certaines de ces mutations provoquent une hyperproduction d’enzymes, jusqu’à cent fois plus
que normalement, avec une baisse de l’affinité pour la triméthopnme (Flensburg, J. and
Skold, 0. 1987). Un autre type de mutation au niveau du chromosome a été observé. Elle fait
en sorte que la bactérie est totalement dépendante de l’approvisionnement externe en
thymines, ce qui lui confère ainsi une résistance à la trirnéthoprime (King, C. H. et aÏ. 1983).
Distribution et mobilité des gènes de résistance
La majorité des gènes de résistance dhfr sont retrouvés sous forme de cassettes. La probabilité
de retrouver les gènes de résistance à la triméthoprime sur les intégrons de classe 1, qui
contiennent normalement à leur extrémité 3’ le gène de résistance suÏI (résistance aux
suLfamides), est favorisée par l’usage combiné de la triméthoprime-sulfamides comme
traitement. Ceci exerce une sélection pour la résistance à ces deux agents antimicrobiens.
L’émergence des gènes de résistance à la triméthoprime parmi les bactéries pathogènes
découle de la facilité de recrutement de ces gènes par des événements de transfert entre les
bactéries. Le gène le plus souvent retrouvé parmi les bactéries à Gram-négatif semble être
dÏfrI. Le fait qu’on le retrouve, dans la majorité des cas, sous forme de cassettes intégrées
dans les intégrons de classe 1 et de classe 2, peut expliquer son émergence parmi ces bactéries
(Skold, 0. 2001). L’intégron de classe 2 est retrouvé sur le transposon Tn7. Celui-ci s’insère
facilement et à fréquence élevée dans le chromosome de plusieurs espèces bactériennes
(Skold, 0. 2001).
Un autre type de transposon a été observé chez des isolats de porcs (Jansson, C. and Skold, 0.
1991). Le gène dÏfrIX a été retrouvé chez plusieurs isolats E. cou provenant de porcs, mais
très rarement dans des isolats provenant d’humains (Jansson, C. et al. 1992). Ce gène semble
être porté par un transposon de type Tn3 en plus de s’apparenter au transposon Tn5393
d’Erwinia spp.
Utilisation de la triméthoprime
La triméthoprime est un agent antimicrobien utilisé le plus souvent en combinaison avec le
sulfaméthoxazole dans le traitement d’infections causées par des bactéries à Gram-négatif.
38
Par contre, récemment, une augmentation de la résistance chez E. cou aux triméthoprime
sulfaméthoxazole, utilisé en premier recourt contre les UTI, a été remarquée (Kahlmeter, G.
2000). En 1970, la triméthoprime était employée seule sous forme prophylactique, puis
comme traitement des infections urinaires sévères (Huovinen, P. et al. 1995). En plus de son
utilisation pour le traitement des infections urinaires, la combinaison trimethoprime
sulfamides peut être utile contre les shigelloses ainsi que contre certaines infections
respiratoires et à staphylocoques (Huovinen, P. et al. 1995).
6.6 Les sulfamides
Les premiers agents antimicrobiens qui ont été utilisés pour leurs propriétés thérapeutiques
sont les sulfamides. Des tests d’efficacité de ces agents ont été faits en 1932 sur des souris
infectées par Streptococcus pyogenes dans la cavité abdominale. Cette expérience a été
publiée en 1935 par Gerhard Domagk.
Les sulfamides sont des agents antimicrobiens de nature synthétique qui possèdent une
analogie de structure avec l’acide p-aminobenzoïque (PABA) impliqué dans la biosynthèse de
l’acide folique (Skold, 0. 2001). Les sulfamides sont en compétition avec l’enzyme la
dihydroptéroate synthase (DHPS) qui catalyse la dernière réaction menant à l’obtention de
l’acide dihydrofolique et qui implique la condensation du PABA et du 7,8-dihydro-6-
hydroxyméthylpterine-pyrophosphate en acide dihydroptéroïque (Skold, 0. 2001). Ces agents
ne peuvent donc pas attaquer les cellules eucaryotes, car elles ne produisent pas d’acide
folique. Les sulfamides ont donc une action bactériostatique très sélective. Puisque les
sulfamides sont des agents antimicrobiens synthétiques, une résistance bactérienne à cet agent
n’avait pas été considérée comme étant possible. Toutefois, la résistance aux sulfamides est
apparue chez les bactéries pathogènes un peu après son introduction en milieu hospitalier vers
les années 40 (Skold, 0. 2001).
Mécanisme d’action de la résistance
Les plasmides qui amènent la résistance aux sulfamides portent des variants des gènes codant
pour l’enzyme dihydroptéroate synthase chromosomiques. Deux gènes codant pour une
39
dihydroptéroate synthase résistante aux sulfamides sont connus, sulI et sutil (Sundstrom, L. et
aï. 198$; Radstrom, P. and Swedberg, G. 1988).
Les deux enzymes codées par les gènes de résistance ont des limitations structurales
importantes qui empêchent une bonne liaison avec l’inhibiteur, les sulfamides, mais ne les
empêchent pas de se lier à leur substrat, le PABA. Bien qu’elles résistent à des concentrations
très fortes de sulfamides, les dihydroptéroate synthase résistantes produites par les gènes sulI
et sulil montrent toutes les deux un Km très faible (0.6iM) pour le PABA, (Skold, 0. 2001).
De plus, l’enzyme du gène sulli semble avoir une acuité particulière au PABA qui permet sa
distinction de l’inhibiteur (Skold, 0. 2001).
Des mutations chromosomiques ont été rapportées, ce qui peut éventuellement affecter la
sensibilité de la souche aux sulfamides. On note, entre autre, un changement dans le gènefolP
codant pour la dihydroptéroate synthase (Skold, 0. 2000).
Distribution et mobilité de la résistance
Le gène sulI est souvent retrouvé avec d’autres gènes de résistance sur l’intégron de classe 1,
tandis que sulil est plutôt localisé sur un petit plasmide de type incQ (R$F1010) ou pBPl
(van Treeck, U. et al. 1981). Ces gènes sont habituellement retrouvés à la même fréquence
dans les isolats cliniques à Gram-négatif (Radstrom, P. et al. 1991). L’origine de ces gènes
n’est pas connue. Chez les bactéries à Gram-négatif, seulement deux plasmides portant ces
deux gènes, sulI et sulIL ont été décrits (Skold, 0. 2001). Probablement que ceci est dû au
fait, qu’en plus d’avoir un Km très faible pour le PABA, ces deux gènes sont localisés sur des
véhicules de dissémination très efficaces. Les deux enzymes montrent 57% d’identité en
acides aminés (Huovinen, P. et aï. 1995).
Utilisation des sulfamides
Les sulfamides sont des agents antimicrobiens à large spectre d’action incluant la plupart des
bactéries pathogènes causant des UTI (E. cou, et autres membres des entérobactéries), les
pathogènes du système respiratoire (Streptococcus pneurnoniae et HearnophiÏus inJluenzae),
les pathogènes de la peau (Staphylococcus aureus), en plus que certains pathogènes
40
entériques (E. cou, Shigella spp.) (Huovinen, P. et al. 1995). À cause de l’éventail des
possibilités d’indications cliniques, les sulfamides sont souvent utilisés en combinaison avec
la triméthoprime. De plus, les deux composés sont relativement peu coûteux, ce qui permet
leur utilisation dans les pays en voie de développement.
41
7.0 Contrôle de la résïstance antimicrobienne
La surveillance de la résistance bactérienne aux agents antimicrobiens est nécessaire afin
limiter l’apparition de celle-ci. L’isolement physique, le contrôle de l’infection ou une
meilleure hygiène, ainsi que de choisir l’agent antimicrobien à utiliser dans le traitement de
l’infection selon le résultat de l’antibiogramme, sont des moyens permettant de limiter
l’émergence de la résistance bactérienne. De plus, la caractérisation de la résistance
bactérienne permet de procéder avec efficacité.
La caractérisation de la résistance aux agents antimicrobiens chez les bactéries peut se faire
par différents moyens. Présentement, différents tests de microbiologie traditionnels tels que la
diffusion des disques en gélose, la dilution en bouillon, la dilution en agar et le gradient de
diffusion, sont utilisés afin de déterminer le phénotype de résistance aux agents antimicrobiens
(Aarts, H. J. et al. 2001). Les bactéries doivent être isolées et mises en culture avant ces tests. La
détection des gènes de résistance aux agents antimicrobiens peut être accomplie par une
amplification PCR dont la cible est l’ADN et où, le résultat est visualisé sur un gel d’agarose.
L’hybridation avec sondes radioactives sur filtre est aussi une méthode permettant de détenuiner
la présence de certains gènes de résistance et ainsi, permettre la caractérisation génotypique de la
résistance bactérienne. La majorité de ces tests demandent plusieurs étapes, les manipulations
peuvent être complexes et peu reproductibles ou demandent beaucoup de temps (Aarts, H. J. et
al. 2001).
Dans certains cas, la détermination du phénotype de résistance des bactéries peut ne pas être
suffisante. L’identification des gènes de résistance aux agents antimicrobiens, dans ces cas,
permet une analyse plus complète. À ce niveau, les méthodes génétiques interviennent et
deveniennent indispensables lors d’études épidémiologiques.
n
SECTION II — Méthodologie et résultats
Article I
Characterization and detection of antimicrobial resistance genes in enterotoxigenic
Escherichia cou 0149: K91 isolated from pigs over a 23-year period
(Accepté par Antimicrobial Agents and Chemotherapy, publication prévue en octobre 2003)
44
Characterization and detection of antimicrobial resistance genes in enterotoxigenic
Escherichia cou 0149: K91 isolated from pigs over a 23-year period
Running titie
Characterization of antimicrobial resistance in ETEC
Christine Maynard’, John M. Fairbrother’, Sadjia Bekal2, François Sanschagrin3, Roger C.
Levesque3, Roland Brousseau2, Luke Masson2, Serge Larivière’ and Josée Harell*
‘Département de Pathologie et Microbiologie, Faculté de Médecine Vétérinaire, Université de
Montréal, Saint-Hyacinthe, Québec, Canada, J2S 7C6, 2Bioteclmology Research Institute,
Montréal, Québec, Canada, H4P 2R2, Pavillon C.- E. Marchand, Université Lavai, Québec,
Québec, Canada, G1K 7P4,
*Conesponding author. Mailing address: Département de Pathologie et Microbiologie,
Faculté de Médecine Vétérinaire, Université de Montréal, 3200, rue Sicotte, C. P. 5000, Saint
Hyacinthe, Québec, Canada, J2S 7C6. Phone: 450 773 $521 ext. $233. Fax 450 778 $10$. E-
mail:
45
ABSTRACT
A total of 112 Escherichia cou 0149: K91 strains isolated from pigs with diarrhea in
Quebec between 1978 and 2000 were characterized for their genotypic antimicrobial
resistance profile. Among the ten antimicrobial agents tested, resistance was most frequent for
tetracycline and sulfonamides, but absent with cefotaxime and ceftiofur. An increase in the
number of isolates resistant to at least three antimicrobials was observed over time. The
distribution of 28 resistance genes covering six antimicrobial families (bcta-lactams,
aminoglycosides, chloramphenicol, tetracycline, trimethoprim and sulfonamides) was
assessed by colony hybridization. Significant differences in the distribution of tetracycline
(tet(A), tet(B), tet(C)), trirnethoprim (dIfrI, dflfrV, dJfrXJII), and sulfonamide (sulI, suUI)
resistance genes were observed in the study period (1978-2000). Sixty percent of the isolates
possessed a class 1 integron illustrating the importance of integrons in the epidemiology of
antibiotic resistance in E. cou isolated from pigs. Amplification of the integron’s variable
region resulted in four distinct fragment sizes of 1, 1.3, 1.6, 1.8-kb with the 1.6 and 1.8-kb
fragments only appearing in the last half of the smdy period. Examination of linkages among
the different resistance genes showed a variety of positive and negative associations.
Association analysis of isolates divided into two groups between 1978-1989 and 1990-2000,
revealed the appearance of new positive resistance gene associations. Our genotypic
resistance analyses of ETEC isolates from pigs clearly indicate that the many of the antibiotic
resistance genes behind phenotypic resistance are flot static but rather in a state of flux driven
by various selection forces such as the use of specific antimicrobials.
46
INTRODUCTION
Enterotoxigenic Escherichia cou (ETEC) is an important swine pathogen causing
dianhea in newbom and postweaning pigs. ETEC of serotype 0149: K91 has been found
more frequently in recent years and is the predominant serotype universally associated with
dianhea in pigs (3, 19, 31). ETEC are also found associated with several other serogroups,
08, 09, 020, 0101, 013$ and 0141 (15). Despite this limited number of serogroups, ETEC
shows diversity of genetic background (3$). With the worldwide progressive increase in
antimicrobial resistance of E. cou isolates, treatment of postweaning diarrhea has become
increasingly difficult. In Quebec, aminopenicillins, chlortetracycline and trirnethoprim
sulfonamides and to a less extent, the aminoglycosides neomycin and apramycin, are the
usual antimicrobials used to treat diarrhea; however, with the appearance of increased
antimicrobial resistance, the third generation cephalosporin, ceftiofur, is seeing increased
usage (3).
The dissemination of antibiotic resistance genes among bacterial strains is an
increasing problem in infectious diseases. Many antibiotic resistance genes are located on
plasmids andlor on transposons, enabling their transfer among a variety of bacterial species.
In recent years, another mechanism of resistance gene dissemination has been discovered,
involving a DNA element that mediates the integration of resistance genes by a site-specific
recombinational mechanism. This newly recognized DNA element, called an integron, is
found either as part of a transposon within the Tn2] family or independently on several
groups of broad-host-range plasmids. Class 1 integrons possess two consewed segments
separated by a variable region (VR) which includes integrated antibiotic resistance genes or
cassettes of unknown function (45). The 3 ‘-conserved segment contains the qacEAl and sulI
genes and an open readmg frame (ORF) called orf5. The qacEAl and sulI genes determine
resistance to ethidium bromide and quatemary ammonium compounds and to sulfonamide,
respectively (45).
Hence, a study to analyze the evolution of different antimicrobial resistance
phenotypes in ETEC 0149: K91 strains isolated from clinical cases of dianhea and other
intestinal disorders in piglets from 1978 to 2000 (F. Fontaine, N. Nadeau, S. D’Allaire, S.
Péres and J. M. Fairbrother, unpublished data), revealed an increase in the number of
antimicrobial multiresistant strains in a bacterial population with time. In the present study,
these same ETEC 0149: K9 1 strains were characterized for their genotypic resistance gene
profiles. Our findings show that during the 23-year study period, antimicrobial resistance gene
47
distribution among E. col! 0149: K91 isoistes is dynamic aM flic obscncd increascs inphcnotypic resistance conelated wiffi an increase in multigene resisance.
48
MATERIALS AND METHODS
Bacterial strains and growth conditions. Numerous E. cou isolates from diseased
animais are sent to The E. cou Laboratory at the Faculté de médecine vétérinaire, at Saint-
Hyacinthe, for serotyping and virulence factor determination (11). For our study, we seiected
ail the ETEC isolates in this collection that came from pigs with diarrhea, belonged to
serogroup 0149: K91, contained virulence factors typicaliy associated with the ETEC
pathotype (e.g. sta, stb, F4, Lt, etc) and were resistant to one of the antimicrobial for which
resistance genes were tested. In order to minimize any selection bias, only one isolate per
farm was selected. The farms from which the isolates were obtained wcre distributed
throughout the various regions of the Province de Quebec. After exciuding 21 isolates which
were susceptible to ail antimicrobials tested, the remaining 112 ETEC which were obtained
over a 23-year period, were divided into four groups based on the year of their isolation: 1978
to 1984, 1985 to 1989, 1990 to 1994 and 1995 to 2000. The reference strain, Escherichia cou
ATCC® 25922, was used to test each lot of antimicrobiai agents by the dise diffusion method.
The bacteriai strains, kept at -80°C in tryptic soy broth (TSB) medium containing 10%
glycerol, were cuitured on tryptic soy agar (TSA) supplemented with 5% (v/v) sheep blood.
The 2$ strains used as positive controis and templates for DNA amplification were
obtained from different laboratories (Table 1). These strains were kept at -80°C as frozen
stocks in tryptic soy broth (TSB) medium containing 10% glycerol (v/v) and were propagated
on Luria Bertani (LB) broth or agar containing one of the following antirnicrobial agents at
the appropriate concentrations: ampicillin (50 jiglml), gentamicin (30 .ig/ml), kanamycin (50
jg!ml), tetracycline (10 tgIml), chioramphenicol (10 tgImi), trimethoprim (10 jig/mi);
suifamethazine (200 tg/ml).
Antimicrobial susceptibility testing. Antimicrobial susceptibiiity testing was carried
out by the dise diffusion method according to the recommendations reported by the National
Comrnittee for Clinical Laboratory Standards (NCCLS) (33). As recommended by the
NCCLS, Muelier-Hinton (MH) agar batches used as the culture medium were tested. The
antimicrobiai agent dises used in this study were: beta-lactams: ampicillin (10 jig), ceftiofur
(30 jig), cefotaxime (30 11g); aminoglycosides: gentamicin (10 kg), kanarnycin (30 jig),
neomycin (30 j.cg); tetracycline (30 jig); chloramphenicol (30 g); trimethoprim (5 jtg);
suifarninoxazole (250 g) (BBL, Bristol, CT, USA.).
The zone diameters were interpreted using the NCCLS recommendations except for
neomycin, for which the breakpoints used were those recommended by the manufacturer. In
49
the case of trimethoprim, interpretation of the zone diameters was according to the method
used for enterobacteria obtained from the unnary tract in the NCCLS recommendations.
PCR primers and amplification. Resistance gene primers were designed using the
GCG software program “Prime” (Genetics Computer Group, Madison, WI, USA).
Oligonucleotide primers were synthesized with a model DNA Synthesizer (BioCorp Inc.,
Ontario, Canada). The PCR primers, their amplified product size as well as the references for
the corresponding strains used as amplification templates are listed in Table 1. The class 1
integron is characterized by the qacEAl and suÏI genes at its 3’-conserved segment (41).
Primers located at the 3’-conserved segment were used as described by Sandvang et al (47) to
investigate the presence of the class I integron (Table 1).
Amplifications were performed on 5 pi of supernatant from bacterial preparations
boiled for 10 min (8).The PCR reaction mix (total 50 tI) included 29.6 Jtl H20, 5.0 pi of IOX
PCR buffer (Amersham Pharmacia Biotech Inc., Piscataway, NJ, USA), 5.0 pi of dNTP 2
mM, 1 U of Taq polymerase (Amersham Pharmacia Biotech Inc.) and 25 pmol of each
primer. DNA amplification was carried out in a GeneAmp® PCR system 9700 (Perkin-Elmer,
Foster City, CA, USA) using the following conditions: 5 min at 94°C followed by 30 cycles
of 94°C for 30 sec, 50°C for 30 sec and 72°C for 1,5 min. A sample of 3 l of the PCR
product was verified for size and purity by gel electrophoresis (1.2 % (w/v) agarose in iX
TAE buffer).
Ail isolates that contained the 3’-conserved segment of the class I integron were
further investigated by another PCR amplification of a variable region (VR) withing the
integron using the following primers described by Sandvang et al (27) (Table 1).
Amplification conditions for these primers were as follows: 5 min at 94°C followed by 35
cycles of 94°C for I mm, 50°C for 1 min and 72°C for 5 min.
UNA sequencing. Amplified products were purified using the QlAquick gel
extraction kit (QIAgen Inc., Mississauga, Ontario, Canada). Their sequences were confirmed
with the dRhodamine Terminator Cycle Sequencing Ready reaction Kit using an Applied
Biosystem 377 DNA sequencer (Applied Biosystems, Foster City, CA, USA). Sequence
comparisons were submitted to the National Center for Biotechnology Information (Bethesda,
Md) for BLAST searches in GenBank. Multiple DNA alignments were perfomied by using
the CLUSTALW program (32).
50
Colony hybridization. The amplicons were Iabelled with ŒP32CTP using the DNA
labelling beads kit (Arnersham Pharmacia Biotech Inc.). Colony hybridizations were
performed as previously described (11).
Statistical methods. Comparisons of association between resistance genes in the total
population were made by using Pearson chi-square’s exact test (SAS, version 8.2, CARY,
N.C.). The statistical significance was f < 0.05. An association between two genes can be
positive indicating that the genes are found together or negative indicating that the genes are
not found together.
51
RESULIS
Antimicrobial resistance phenotype charaeteristics. Almost ail (93%) of the 112isolates were resistant to tetracycline, and a similar number (9 1%) were resistant tosulfonarnides. The percentage of resistance to ampicillin, neomycin, kanamycin,chloramphenicol and to trimethoprim ranged from 21 to 38%, whereas only 14% isolateswere resistant to gentamicin (Table 2). Finally, none of the tested isolates was resistant tocefotaxime or ceftioftir. Some isolates demonstrated a zone of inhibition at the limit (zonebetween resistance and susceptibility) to aminoglycosides, tetracycline and sulfonamides. Tofacilitate our analysis, these isolates were considercd as resistant in our study.
As shown in Table 2, the percentage of multiresistance increased over tirne for certainantimicrobials. Between 1978 and 19$4, the percentage of isolates resistant to neomycin,kanamycin, tetracycline and sulfonamides was already quite high. The resistance toampicillin, gentamicin, chloramphenicol and trimethopnm showed the greatest increases inthe last pcriod (1995-2000). The percentage of strains resistant to at least three antimicrobialsisolated in the periods between 1978-1984, 1985-1989, 1990-1994 and 1995-2000, were 44%,31%, 68% and 83% respectively (data flot shown).
Distribution of resistance genes in 0149: K91 ETEC isolates. The choice of theresistance genes to be studied was based on their relative importance as observed in resistantE. cou isolates (5, 6, 13, 20, 21, 27, 35). Therefore, 28 genes coding for antimicrobialresistance, belonging to six antimicrobial families (beta-lactams, aminoglycosides,tetracycline, phenicols, trimethoprim and sulfonarnides), were chosen to determine theirdistribution in 0149: K91 ETEC isolates from pigs with diarrhea (Table 1).
(I) Beta-Jactams. The DNA hybridization probes blanM and blaSllv detect ail theknown variants within the conesponding bÏaTEM and blasHv gene families on colonyhybridization. The probe blaoxp7 detects bÏaoxp. variants sucli as blaoxio to blaOXAl4,blaoxp6 to blaO)(Jl9, blaox2s, bÏaoXA3t, bÏa035 (80% to 96% of similarity) and bÏapsE2(96.3% of similarity). To further discriminate variants among the blaox7 hybridizationpositive isolates, a PCR-amplification was undertaken using the blaox7 specific primers. Ailthe 112 smdied isolates possessed at least one of the tested resistance genes.
Eighty-six percent of the ampicillin-resistant isolates and ail the blasHv-positiveisolates were bÏaTEM-positive. The blaoxjj gene was only found in 25% of ampicillinresistant isolates isolated during the third period (1990-1994). None ofthe tested isolates were
positive for the blacTxM3, bÏapsE4 or bÏaoxp7 probes. One of the 69 isolates susceptible to
52
beta-lactams hybridized with blaTEM and four isolates resistant to ampiciilin did not possess
any of the tested beta-lactam genes.
(ii) Aminoglycosides. 0f the five tested aminoglycoside resistance genes, only
aph(3 9-la, aph(3 9-lia and aac(3,) -IV were found among the resistant isolates (Table 3). The
aph(3 9-la and aph(3 9-lia genes, encoding a kanamycin and a neomycin resistance
phenotype, were found in 87% and 15% of the neomycin-resistant isolates and in 79% and
19% of the kanamycin-resistant isolates, respectively. The relative importance of aph(3 9-la
and aph(3 9-lia varied throughout the studied periods (1978-2000). Two isolates harbouring
the aph(3 9-la gene were susceptible to kanamycin and neomycin. Only one gentamicin
resistance gene, aac(3)-IV, was found among the isolates. Seventy-five percent of the
gentamicin-resistant isolates and two gentamicin susceptible isolates had this gene.
(iii) Tetracycline. 0f the six targeted tetracycline resistance genes, only the tet(A),
tet(B) and tet(C) genes were detected. The tet(B) gene was found in 80% of the tetracycline
resistant isolates and was by far the most frequentiy observed gene found during the first three
periods. Only 25% tetracycline-resistant isolates were positive for the tet(A) and tet(C) probes
over the whole study period. These two genes, Iess prevalent in the first periods (< 15%),
appeared even slightly more freqently (54%) than tet(B) (5 1%) between 1995 and 2000
(Table 3). An association between the tet(A) and tet(C) genes was observed in ail the tet(A)
positive isoiates and oniy 5% of the tet-positive isoiates possessed ail three tetracycline
resistance genes. No link between these genes and phenotype was observed in four isolates:
two of the 8 isolates susceptible to tetracycline were tet(B)-positive and two of the 104
tetracyciine-resistant isolates were negative for the six tested tet genes.
(iv) Phenicols. Only the chloramphenicol resistance genes catI andftoR were detected
among the tested isolates. On one hand, the gene catI was found in 79% of the
chloramphenicol-resistant isolates. The foR gene was flot detected in chloramphenicol
resistant isolates. Three chioramphenicol susceptible isolates were positive for the catI probe
and one strain isolated in 1989 was foR-positive but was chloramphenicol susceptible. Four
of the 24 chloramphenicol-resistant isolates did flot hybridize with the tested cat genes.
(y) Trimethoprim. In our study, the trimethoprim resistance phenotype was found to
be associated with the presence of three genes, dÏifrI, dÏfrV and dflfrXlII. These genes were
flot detected in strains isolated from 1978 to 1984. The genes dÏfrI was detected in strains
isolated in the second period (198 5-1989), whereas dIfrXIII appeared in the third one (1990-
1994). AIl three trimethoprim resistance genes were found among the strains isoÏated in the
53
last period with a predominance of dÏfrV. Among the trimethoprim-resistant isolates, three
were negative for the tested genes whereas one susceptible isolate was found to be dhfrXJII
positive.
(vi) Sulfonamides. Seventy-nine percent of sulfonamide-resistant isolates possessed
the gene sulI as compared to 36% that possessed suÏII (Table 3). Seventeen percent of these
resistant isolates possessed both sulI and sulli. The percentage of the positive isolates to sulI
did not vary significantly during any ofthe periods. However, the percentage ofsuÏll-positive
isolates increased significantly to 56% in the last period. No conelation between the presence
of sulfonamide resistance genes and phenotype was observed for six isolates. four of the 10
sulfonamide-susceptible isolates were sulI-positive, whereas two of the 102 sulfonamide
resistant isolates were found to be sulI- and sutIJ-negative.
Identification of integrons. 0f the 112 isolates, 67 (60%) amplified a PCR fragment
from the class 1 integron 3 ‘-conserved region (Table 4). Among these positive isolates, 84%
were also positive for a sulI probe by colony hybridization (data flot shown). Among the 67
class 1 integron-positive isolates, four distinct amplicons of 1, 1.3, 1.6 and 1.8-kb were
obtained by amplification of the variable region (VR) using the primers Finti and Rintl.
More specifically, more than haif of the isolates (5 5%) amplified a 1.3-kb VR fragment,
followed by 29% amplifying a 1-kb fragment. The remaining 16% of the isolates produced
fragments larger than 1.3-kb. The percentage of isolates exhibiting the larger 1.6 and 1.8-kb
amplified VR fragments increased over time. Before 1990, none of the class 1 integron
positive isolates had a VR fragment larger than 1.3-kb. However, a larger VR fragment
appeared in one (6%) of such isolates in the period 1990-1994, and the frequency of isolates
with larger VR fragment increascd to 50% in the latest period, 1995-2000.
Two representative \TR fragments of each length were sequenced. Both VR fragments
of the same length contained the same gene cassettes with more than 95% of similarity. The
1-kb VR fragment contained the ant(3 ‘9-la (aadA]) gene cassette (accession no. X 12870)
encoding for streptomycin and spectinomycin resistance. The most frequently observed VR
amplified fragments (i.e. 1.3-kb) contained, in addition to the ant(3 ‘9-la gene cassette found
in the 1-kb fragment, an orf named orm (accession no. M86913 and AF140629). The 1.6-kb
VR fragments contained two gene cassettes, dhfrlb, encoding for resistance to trimethoprim,
and ant(3 ‘9-If (aadA6), encoding for resistance to streptomycin and spectinomycin (accession
no. AF393510 and AF140629). finally, the 1.8-kb VR fragments possessed two gene
cassettes, dhfrXII and ant(3 “)-If which confer a resistance to trimethoprim and a resistance to
54
streptomycin and spectinomycin, and one orf named orfF (accession no. Z21672 and
Af284063).
Association between resistance genes. In order to see possible associations between
resistance genes found among our isolates and if the co apparition of some resistance genes
was statistically confirmed, the comparison of association was done using Pearson chi-sqare’s
exact test. Significant associations with respect to the occurrence of individual resistance
genes among the whole collection of ETEC isolates were detected (P <0.05) (Table 5). Some
positive associations were obvious; for example, the association of bÏaTEM and blasHv as well
of tet(A) and tet(C) genes. The blaTEM gene was associated with the aph(3 9-la, aac(3,) -IV,
dhfrV, and catI genes and also, although less strongly, associated with the sulli gene (Table
5). The aac(3)-IV gene showed a positive association with blaTEM and bÏasnv, and also with
aph(3 9-la, dhfrV, and dhfrXll. The aph(3 9-lia gene was positively associated with tet(A) and
tet(C) but negatively associated with tet(B). In contrast, the sulI gene was positively
associated with tet(B) but negatively associated with the tet(A) and tet(C) genes. Although the
class 1 integron was found together with the suÏI gene in 68% of the sutI-positive isolates, no
statistical association was observed between the presence of class 1 integron and sulI.
Probably this no correlation is due to the presence of sulI in integron-negative isolates.
The association analysis was subsequently performed on the collection of isolates
divided into two groups, group 1 representing those strains isolated between 1978 and 1989
and group 2 representing those isolated between 1990 and 2000 (data not shown). In the
period 1990 to 2000, new positive associations were observed, such as blaTEM with bÏasHv,
sutI with tet(A), tet(B), tet(C) and the class 1 integron, and finally catI with the class 1
integron. Some positive associations appeared more closely linked than previously observed
in the whole collection; for example blaTEM with aph(3 9-la, aac(3)-IV and dhfrXfII with
sulil.
55
DISCUSSION
ETEC is the most common bacterial etiologic agent of dianhea in neonatal and
postweaning pigs. Although treatment of enteric E. cou infection in swine commonly includes
the use of antimicrobials (2, 10), relatively few studies have been directed towards the
characterization of genotypic resistance profiles of E. cou strains isolated from animais. One
study characterized the tetracycline and sulfonamides resistance gene profiles of E. cou
isolated from animais and humans (17) while a second (27) characterized the aminoglycoside
resistance gene profiles of porcine and bovine E. cou isolates. In the present study, ETEC
0149: K91 strains isolated from pigs between 1978 and 2000 were characterized for their
acquired resistance gene profiles against ten different antimicrobial agents.
In this study, phenotypic resistance was overestimated because only isolates showing
initial resistance to one antimicrobial were further anaiyzed. Nevertheless, it remains that the
phenotypic resistance observations reported here reflect the general trend observed with E.
cou isolated from pigs (3, 19, 31). Although we did not test for all the antimicrobials which
could inhibit growth of E. cou (other aminoglycosides, quinolones/fluoroquinolones, colistin)
most of the resistant isolates were resistant to more than one antimicrobial, especially during
the last period (1995-2000). No resistance to cephalosporins (ceftiofur and cefotaxime) was
observed.
The resistance to tetracycline and sulfonamide antimicrobials observed in most of our
ETEC isolates was also noted by other groups in pig isolates (17, 31). It is not surprising that
the level of resistance of isolates to these antimicrobials was so high, as these antimicrobials
have been and are stiil being used as growth promoters in Canada and USA, in disease
prevention and in therapy in swine production (12). Among the tetracycline-resistant isolates,
the tet(B) gene was largely predominant untii 1994 when two other closely associated
tetracycline resistance genes, tet(A) and tet(C), became dominant during the last study period
(1995-2000). Another study done on pigs from three herds with different histories of
antimicrobiai exposure produced similar results in that tet(B) was predominant in two of the
herds exposed to antimicrobials and when present, tet(A) and tet(C) were also found together
(18). In contrast, a recent study by Lanz et aÏ (17) showed that the tet(A) gene alone was the
most prevalent tet gene in E. cou isolates from pig with dianhea or enterotoxaemia. The mode
of action as well as the specificity of certain antimicrobial enzymes could exert positive
selection pressure and contribute to the emergence of new genes over time. For example, ciass
A, B and C tetracycline-resistant determinants are efflux pumps with different specificities.
56
Most of the efflux proteins confer resistance to tetracycline but not to minocyline or
glycylcycline antimicrobial groups. In contrast, the tet(B) gene encodes for an efflux protein
which confers resistance to tetracycline, doxycycline and minocycline but flot glycylcycline
(24). These specificities correlate with the emergence of the diverse distribution of different
tetracycline resistance genes over time. It is not known if such a selective effect exists
between the commonly used tetracyclines in swine production, i.e. oxytetracycline and
chlortetracycline. Similarly, the sutI gene was more predominant than sulli in strains isolated
during the three first periods, whereas, the sulil gene appeared as frequently as sulI in the iast
period (1995-2000). The predominance of sulI among pig isolates was also observed by Lanz
et aÏ (17). The sulI and sulli genes encode dihydropteroate synthase enzymes with different
sensitivities (K), even if the two enzymes show the same low Km values (0.6jiM) for PABA
which is implicated in the bacterial folic acid biosynthesis. The enzyme encoded by suÏII
discems the normal PABA substrate from the inhibitor, the sulfonamides (29).
In spite ofa ban on the use ofchloramphenicol in food animais in Canada since 1980
(9), an increase of chloramphenicol resistance in ETEC isolates was observed. The
persistence or increase of chloramphenicol resistance in E. cou has been observed by others in
swine (3, 17, 31) and from other animal species (34). Resistance to choramphenicol was
closely associated with the presence of the catI gene. In a study done by Bischoff et aÏ (7),
only four out of 48 chloramphenicol-resistant isolates were harboring a cadI gene; the
relatively unknown gene, cmÏA, was responsible for the resistance of the other isolates.
finally, we detected the foR gene in only one isolate. Other studies report the presence of
foR in a large number of E. cou isolated from chickens and cattie (7, 16, 35).
Interestingly, 35% of the isolates were resistant to neomycin and 38% to kanamycin
even though kanamycin is not used in the Canadian swine industry. It is likely the resuit of
cross-resistance by most of the aminoglycosides resistance genes (34). Our study showed that
certain antimicrobial resistance genes are more prevalent than others and that this incidence
changed over time within the ETEC. The blaTEM genes were widely distributed whereas the
blaoxj. genes appeared infrequently. Among the genes tested in our ETEC isolates, only the
aminoglycoside resistance genes revealed limited diversity with most prevalent genes being
aph(3 9-la and aph(3 9-lia. These genes, which are also responsible for cross resistance to
different aminoglycosides (neomycin and kanamycin), did not vary in frequency over the
duration of our study. In contrast, a Danish publication showed a varying frequency among
57
the aminoglycoside resistance genes with ant(2 ‘9-la and aac(3)-lia being the most prevalent
among pig isolates (27).
One clear observation arising from our study is that the number and diversity of genes
driving the phenotypic resistance is dynamic. During the 23-year period, some genes or
association of genes appeared whereas other genes became more rare (Table 5). The
association of the blasHV gene with blaTEM could be explained by the fact that the bÏasFiv gene
might have been acquired in isolates harboring blaTEM resulting in an increase of resistance to
beta-lactams (4). We observed that the tet(A) and tet(C) genes were aiways found together.
During the three first periods, tet(B) was the most prevalent tetracycline resistance gene.
Incompatibility of plasmids carrying tetracycline resistance determinants could explain the
existence ofthe negative associations between tet(A)Itet(C) and tet(B) (15).
Resistance genes are associated with mobile DNA such as plasmids, transposons and
integrons which facilitates resistance gene distribution (14, 30). Most of the isolates (62%)
possessed a class 1 integron. Because integrons are characterized by their integration ofmany
different gene cassettes between insertion sites in the VR, site-specific insertion represents
another mechanism driving the evolution of plasmids and transposons of gram-negative
bacteria. In our E. cou collection, most class 1 integrons contained VRs of 1.3 and l-kb. The
VR of integrons found in Shiga toxin-producing E. cou (STEC) isolates showed similar sizes
(36). Interestingly, an increase in VR length was observed in E. cou strains isolated during the
last period, a phenomenon also observed by Schmitz et al (28). They showed that the VR of
the class 1 integrons in human E. cou strains isolated in 1993 ranged from 0.65 to 1.8-kb and
those isolated in 1999, from 0.75 to 3-kb, suggesting an accumulation of inserted gene
cassettes into the class 1 integron among the ETEC isolates, due to selection by
antimicrobials. In this study, most of the tested isolates which possessed the same length of
VR appear to have acquired the same gene cassettes into their integron. However, genes
responsible for the resistance to beta-lactams, tetracycline and chloramphenicol were not
associated with class 1 integrons.
Multiple cassette insertions and more than 40 distinct cassettes have been identified
among integrons (25). Among the class 1 integrons of the isolates, four different sizes of VR
were detected (Table 4). Representatives of these VRs were sequenced. As in other studies,
our study shows that the different gene cassettes of the integron VRs included genes encoding
for aminoglycosides and/or trimethoprim resistance, which are the most frequently described
antimicrobial resistance gene cassettes (1, 23, 26).
58
In the last period (1995-2000), we observed an increase in the number of isolates
having multiple resistance genes as well as the appearance of resistance genes such as aac(3)-
IÏ’Ç dhfrv, and sulli . These observed increases in resistance to the different antimicrobials are
presumably due to increased selection pressure resulting from changed management styles,
for example, the early medicated weaning of piglets introduced in early nineties. The direct
use of antimicrobials can drive the co-selection of resistance genes. For example, the use of
injectable oxytetracycline in cattle receiving in-feed chlortetracycline was associated with an
increase in the incidence of resistance to chloramphenicol and sulfisoxazole (22). In our
study, the association of aac(3)-IV, catI and dhfrV, encoding for resistance to gentamicin,
chloramphenicol and trimethoprim, respectively, was observed and the incidence of these
genes increased in a similar distribution over time. This suggests that the increased use of
gentamicin or sulfa-trimethoprim in pig production could have co-selected for the resistance
to chloramphenicol thus explaining the increase in chloramphenicol-resistant isolates, even
though this antimicrobial agent has not been used in swine production since 1980.
In conclusion, our genotypic resistance analyses of ETEC isolates show that the genes
behind phenotypic resistance are not static but rather in a state of flux driven by various
selection forces such as the use of specific antimicrobials. Often, more than one gene was
associated with a given phenotypic resistance. A different distribution of resistance genes was
observed over time: an increase in multigene resistance correlating with the observed
phenotypic multiresistance among ETEC 0149: K91 strains. The difference obsewed with
studies in other countries suggest that relative resistance genes frequencies not only vary over
time within a population but also between populations of different geographical origins. This
smdy reinforces the necessity of using genotypic resistance analyses in future epidemiology
studies.
59
ACKNOWLEDGMENIS
We are very grateful to Guy Beauchamp for statistical analysis and to Kim Messier for
phenotypic resistance characterization of isolates. We appreciate the technical assistance
provided by the members ofthe Groupe de Recherche sur les Maladies du Porc (GREMIP).
This work was supported by the Natural Sciences and Engineering Research Council
of Canada (NSERC), being part of a research work unit of the Canadian Research Network on
Bacterial Pathogens of Swine grant 22M 55 and by the Fédération des Producteurs de Porcs
du Québec (FPPQ).
60
REFERENCES
1. Bass, L., C. A. Liebert, M. D. Lee, A. O. Summers, D. G. White, S. G. Ihayer,
and J. J. Maurer. 1999. Incidence and characterization of integrons, genetic elements
mediating multiple-drug resistance, in avian Escherichia cou. Antimicrob Agents
Chemother 43:2925-9.
2. Bertchfnger, H. U., and J. M. Fairbrother. 1999. Escherichia cou Infections, p.
431-46$. In B. E. Straw, S. DAllaire, W. L. Mengeling, and D. J. Taylor (ed.),
Diseases of swine, $ ed. Iowa State University Press, Ames.
3. Bischoff, K. M., D. G. White, P. F. McDermott, S. Zhao, S. Gaines, J. J. Maurer,
and D. J. Nisbet. 2002. Characterization of chloramphenicol resistance in beta
hemolytic Escherichia cou associated with diarrhea in neonatal swine. J Clin
Microbiol 40:389-94.
4. Bradford, P. A., C. Urban, A. Jaiswal, N. Mariano, B. A. Rasmussen, S. J. Projan,
J. J. Rahal, and K. Bush. 1995. SHV-7, a nove! cefotaxime-hydro!yzing beta
lactamase, identified in Escherichia cou isolates from hospitalized nursing home
patients. Antimicrob Agents Chemother 39:899-905.
5. Bush, K., G. A. Jacoby, and A. A. Medeiros. 1995. A functiona! classification
scheme for beta-lactamases and its cone!ation with molecu!ar structure. Antimicrob
Agents Chemother 39:1211-33.
6. Chopra, I., and M. Roberts. 2001. Tetracyc!ine antibiotics: mode of action,
applications, molecu!ar biology, and epidemio!ogy of bacteria! resistance. Microbio!
Mol Biol Rev 65:232-60 ; second page, table of contents.
7. Cloeckaert, A., S. Baucheron, G. Flaujac, S. Schwarz, C. Kehrenberg, J. L.
Martel, and E. Chaslus-Dancla. 2000. Plasmid-mediated florfenico! resistance
encoded by thefloR gene in Escherichia cou iso!ated from catt!e. Antimicrob Agents
Chemother 44:2858-60.
8. Daigle, F., J. Harel, J. M. Fairbrother, and P. Lebel. 1994. Expression and
detection of pap-, sfa-, and aJa-encoded fimbria! adhesin systems among
uropathogenic Escherichia cou. Can J Microbiol 40:286-9 1.
9. Gïlmore, A. 1986. Ch!oramphenicol and the politics of hea!th. Cmaj 134:423, 426-8,
433-5.
61
10. Hampson, D. J. 1994. Postweaning Eseherichia cou diarrhoea in pigs, p. 171-191. hi
C. L. Gyles (ed.), Escherichia cou in domestic animais and humans. CAB
international, Gueiph.
11. Harel, J., H. Lapointe, A. Fallara, L. A. Lortie, M. Bigras-Poulin, S. Lariviere,
and J. M. Fairbrother. 1991. Detection of genes for fimbrial antigens and
enterotoxins associated with Escherichia cou serogroups isolated from pigs with
diarrhea. J Clin Microbiol 29:745-52.
12. Health Canada. 2002. Uses ofantimicrobial drngs in food animais, p. 53-67, Used of
antimicrobiais in food animais in Canada Impact on resistance and human health.
Heaith Products and food branch, Canada.
13. Huovinen, P., L. Sundstrom, G. Swedberg, and O. Skold. 1995. Trimethoprim and
sulfonamide resistance. Antimicrob Agents Chemother 39:279-$9.
14. Jacoby, G. A. 1994. Extrachromosomal resistance in gram-negative organisms: the
evolution ofbeta-iactamase. Trends Microbiol 2:357-60.
15. Joncs, C. S., D. J. Osborne, and J. Staniey. 1992. Enterobacterial tetracyciine
resistance in reiation to piasmid incompatibility. Moi Celi Probes 6:313-7.
16. Keyes, K., C. Hudson, J. J. Maurer, S. Thayer, D. G. White, and M. D. Lee. 2000.
Detection of florfenicol resistance genes in Escherichia cou isolated from sick
chickens. Antimicrob Agents Chemother 44:42 1-4.
17. Lanz, R., P. Kuhnert, and P. Boerlin. 2003. Antimicrobial resistance and resistance
gene determinants in clinical Escherichia cou from different animai species in
Switzerland. Vet Microbiol 91:73-$4.
18. Lee, C., B. E. Langlois, and K. A. Dawson. 1993. Detection of tetracyciine
resistance determinants in pig isolates from three herds with different histories of
antimicrobiai agent exposure. Appi Environ Microbiol 59:1467-72.
19. Mathew, A. G., A. M. Saxton, W. G. Upchurch, and S. E. Chattin. 1999. Multipie
antibiotic resistance pattems of EscÏzerichia cou isoiates from swine farrns. Appi
Environ Microbioi 65:2770-2.
20. Miller, G. H., F. J. Sabateili, R. S. Hare, Y. Glupczynsld, P. Mackey, D. Shiaes, K.
Shimïzu, and K. J. Shaw. 1997. The most frequent aminoglycoside resistance
62
mechanisms--changes with time and geographic area: a reflection of aminoglycoside
usage pattems? Aminoglycoside Resistance Study Groups. Clin Infect Dis 24 Suppi
1 :$46-62.
21. Murray, I. A., and W. V. Shaw. 1997. O-Acetyltransferases for chloramphenicol and
other natural products. Antimicrob Agents Chemother 41:1-6.
22. O’Connor, A. M., C. Poppe, and S. A. Mcfwen. 2002. Changes in the prevalence of
resistant Escherichia cou in cattie receiving subcutaneously injectable oxytetracycline
in addition to in- feed chlortetracycline compared with caille receiving only in-feed
chlortetracycline. Can J Vet Res 66:145-50.
23. Peters, E. D., M. A. Leverstein-van Hall, A. T. Box, J. Verhoef, and A. C. Finit.
2001. Novel gene cassettes and integrons. Antimicrob Agents Chemother 45:2961-4.
24. Petersen, P. J., N. V. Jacobus, W. J. Weiss, P. E. Sum, and R T. Testa. 1999. In
vitro and in vivo antibacterial activities of a nove! glycylcycline, the 9-t-
butylglycylamido derivative of minocycline (GAR-936). Antimicrob Agents
Chemother 43:738-44.
25. Recchia, G. D., and R. M. Hall. 1997. Origins of the mobile gene cassettes found in
integrons. Trends Microbiol 5:389-94.
26. Roy, P. H. 1997. Dissémination de la résistance aux antibiotiques le génie génétique
à l’oeuvre chez les bactéries. Médecine/Sciences 13:927-933.
27. Sandvang, D., and F. M. Aarestrup. 2000. Characterization of aminoglycoside
resistance genes and class 1 integrons in porcine and bovine gentamicin-resistant
Escherichia cou. Microb Drug Resist 6:19-27.
28. Schmitz, F. J., D. Hafner, R. Geisel, P. Follmann, C. Kirschke, J. Verhoef, K.
Kohrer, and A. C. Fluit. 2001. Increased prevalence of class I integrons in
Escherichia cou, KÏebsieÏla species, and Enterobacter species isolates over a 7-year
period in a German university hospital. J Clin Microbiol 39:3724-6.
29. Skold, 0. 2001. Resistance to trimethoprim and sulfonamides. Vet Res 32:26 1-73.
30. Tenover, F. C., and J. K. Rasheed. 1998. Genetic methods for detecting
antimicrobial and antiviral resistance genes, p. 157$-1592. In P. R. Murray (ed.),
Manual of clinical microbiology, 7 ed, Washington, D.C.
63
31. Teshager, 1., 1. A. Herrero, M. C. Porrero, J. Garde, M. A. Moreno, and L.
Dominguez. 2000. Surveillance of antimicrobial resistance in Escherichia cou strains
isolated from pigs at Spanish slaughterhouses. Int J Antimicrob Agents 15:137-42.
32. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving
the sensitivity of progressive multiple sequence alignment through sequence
weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids
Res 22:4673-80.
33. Watts, J. L., M. M. Chengappa, J. R. Cole, J. M. Cooper, T. J. Inzana, M. R.
Plaunt, T. R. Shryock, C. Thornsberry, R. D. Walker, and C. Wu. 1999.
Performance standards for antimicrobiai disk susceptibility tests for bacteria isolated
from animais; Approved standard document, M31-A ed, vol. 19, Wayne,
Pennsylvania.
34. Werckenthin, C., S. Seidi, J. Riedi, E. Kiossis, G. Wolf, R. Stoila, and O. R.
Kaaden. 2002. Escherichia cou isolates from young calves in Bavaria: in vitro
susceptibilities to 14 anti-microbial agents. J Vet Med B Infect Dis Vet Public Health
49:61-5.
35. White, D. G., C. Hudson, J. J. Maurer, S. Ayers, S. Zhao, M. D. Lee, L. Bolton, T.
Foley, and J. Sherwood. 2000. Characterization of chloramphenicol and florfenicol
resistance in Escherichia cou associated with bovine dianhea. J Clin Microbiol
38:4593-8.
36. Zhao, S., D. G. White, B. Ge, S. Ayers, S. Friedman, L. English, D. Wagner, S.
Gaines, and J. Meng. 2001. Identification and characterization of integron-mediated
antibiotic resistance among Shiga toxin-producing Escherichia cou isolates. Appl
Environ Microbiol 67:1558-64.
DD
D
TA
BL
E1.
PC
Rp
rîm
ers
used
for
anti
mic
robia
lre
sist
ance
gene
and
clas
s1
inte
gro
nam
plif
icat
ions
bÏa
CT
XM
3
ant(
2‘9
-la
(aad
3Y
aac(
3)-
IIa
(aac
C2)
aac(
3)-I
V
aph(
39-
la
(aph
A1)
aphf
39
-lia
(aph
A2)
64
Ant
imic
robi
alR
esis
tanc
e.
Am
plic
onG
enba
nk.
For
war
dP
CR
prim
erse
quen
ce5’
-3’
Rev
erse
PC
Rpr
imer
sequ
ence
5’-3
’So
urce
ofD
NA
fam
ilyge
nesi
ze(b
p)ac
cess
ion
no.
Bet
a-L
acta
ms
bÏa-
1-EM
GA
GT
AT
FC
AA
CA
TF
FtC
GT
AC
CA
AT
GC
TA
AT
CA
GT
GA
$57
AF
3098
24R
.C
.L
eves
que
Am
inog
lyco
side
s
Tet
racy
clin
e
bla
sHv
TC
GC
CT
GT
GT
AT
FA
TC
TC
CC
CG
CA
GA
TA
.AT
CA
CC
AC
AA
TG
76$
AFY
4$85
0R
.C
.L
eves
que
btao
xpi
GC
AG
CG
CC
AG
TG
CA
TC
AA
CC
CG
CA
TC
AA
AT
GC
CA
TA
AG
TG
19$
AJ2
3834
9P
aste
urIn
stit
ute
blao
xp7
AG
UC
TC
TG
CC
GA
AG
CC
TC
TC
AA
CC
CA
AC
CA
AC
CC
591
X75
562
R.
C.
Lev
esqu
e
bla
PS
E4
CT
GC
TC
GT
AT
AG
GT
GT
TC
CT
CG
CA
TC
AT
TT
CG
CT
CT
TC
705
J05
162
R.
C.
Lev
esqu
e
AA
TC
AC
TG
CG
TC
AG
TT
CA
CT
TT
AT
CC
CC
CA
CA
AC
CC
AG
701
X92
506
A.
Hul
etsk
y
TC
CA
GA
AC
CfG
AC
CG
AA
CG
CA
AG
AC
CT
CA
AC
C1T
FF
CC
700
X04
555
R.
C.
Lev
esqu
e
CG
GA
AG
GC
AA
TA
AC
GG
AG
TC
GA
AC
AG
GT
AG
CA
CT
GA
G74
0X
5472
3D
.S
andv
ang
GT
GT
GC
TG
CT
GG
TC
CA
CA
GC
AG
TT
GA
CC
CA
GG
GC
TG
TC
GC
627
X0
13$5
J.H
arel
AT
GG
GC
TC
GC
GA
TA
AT
GT
CC
TC
AC
CG
AG
GC
AG
TrC
CA
T60
0M
l832
9J.
Har
el
GA
AC
AA
GA
TG
GA
TF
GC
AC
GC
GC
TC
FrC
AG
CA
AT
AT
CA
CG
G68
0V
006
1$J.
Har
et
tet(
A)
GT
GA
AA
CC
CA
AC
AT
AC
CC
CG
AA
GG
CA
AG
CA
GG
AT
GT
AG
888
X00
006
J.H
arel
tet(
B)
CC
TfA
TC
AT
GC
CA
GT
CG
CA
CT
GC
CG
Tfl
TT
CG
CC
774
JOl$
30J.
Har
et
tet(
C)
AC
UG
GA
GC
CA
CT
AT
CG
AC
CT
AC
AA
TC
CA
TG
CC
AA
CC
C88
1JO
l749
J.H
arel
tet(
D)
TG
GG
CA
GA
TG
GT
CA
GA
TA
AG
CA
GC
AC
AC
CC
TG
TA
GT
FF
FC
827
X65
876
S.B
.L
evy
tet(
E)
TfA
AT
GG
CA
&C
AG
CC
AG
CT
CC
AT
AC
CC
AT
CC
AT
TC
CA
C$5
3L
0694
0M
.C
.R
ober
ts
tet(
Y)
AC
CG
CA
CT
CA
TG
GT
CT
fCC
AA
GC
AG
CA
AC
AC
AC
$23
Af0
7099
9M
.C
.R
ober
ts
DD
D65
Phe
nico
lsca
tiA
GT
FG
CT
CA
AT
GT
AC
CT
AT
AA
CC
YG
TA
AT
fCA
TT
AA
GC
AT
TC
TG
CC
547
M62
822
J.H
arel
cati
!A
CA
CT
FF
GC
CC
TT
TA
TC
GT
CT
GA
AA
GC
CA
TC
AC
AT
AC
TG
C54
3X
5379
6P
aste
urIn
stitu
te
cati
ilY
FC
GC
CG
TG
AG
CA
TT
TT
GT
CG
GA
IGA
GT
AT
GG
GC
AA
C28
6X
07$4
8I.
A.
Mur
ray
foR
CG
CC
GT
CA
TF
CC
TC
AC
CH
CG
AT
CA
CG
GG
CC
AC
GC
TG
TG
TC
215
Af2
5285
5D
.G
.W
hite
Tri
met
ho
pri
mdh
frI
AA
GA
AT
GG
AG
TrA
TC
GG
GA
AT
GG
GG
TA
AA
ÀA
CT
GG
CC
TA
AA
AT
TG
391
X00
926
J.H
arel
dhfr
VC
TG
CA
AA
AG
CG
AA
AA
AC
GG
AG
CA
AT
AG
TF
AA
TG
TF
TG
AG
CT
AA
AG
432
Xl 2
868
0.
Skôl
d
dhfr
VII
GG
TA
ÀT
GG
CC
CT
GA
TA
TC
CC
TG
TA
GA
TrT
GA
CC
GC
CA
CC
265
X58
425
0.
Skôl
d
dhfr
IXT
CT
AA
AC
AT
GA
TrG
TC
GC
TG
TC
flG
TT
TT
CA
GT
AA
TG
GT
CG
GG
462
X57
730
C.
Wal
len
dhfr
XH
IC
AG
GT
GA
GC
AG
AA
GA
rrY
FT
CC
TC
AA
AG
GT
rTG
AIG
TA
CC
294
Z50
802
P.V
.A
dria
n
Su
lfo
nam
ides
sut!
TfC
GG
CA
TfC
TG
AA
TC
TC
AC
AT
GA
TC
TA
AC
CC
TC
GG
TC
TC
822
X12
869
R.
C.
Lev
esqu
e
sull
iC
GG
CA
TC
GT
CA
AC
AT
AÀ
CC
GT
GT
GC
GG
AT
GA
AG
TC
AG
722
M36
657
J.H
arel
Cla
ssI
inte
gro
nqa
cEM
-su
t]A
TC
GC
AA
TA
Gfl
’GG
CG
AA
GT
GC
AA
GG
CG
GA
I\A
CC
CG
CG
CC
797
X12
870
VR
GG
CA
TC
CA
AG
CA
GC
AA
GC
AA
GC
AG
AC
TT
GA
CC
TG
AT
Var
iabl
eX
l 287
0
Alt
erna
tive
nom
encl
atur
e.
TA
BL
E2.
Tre
nds
inre
sist
ance
ofE
TE
C0149:
K91
isol
ates
wit
htï
me,
asdet
erm
ined
byth
edi
scdi
ffus
ion
met
hod
Num
ber
of
resi
stan
tE
TE
C0
14
9:
K91
isol
ates
(%)a
197$
-198
419
85-1
989
1990
-199
419
95-2
000
AT
M’
(n=
25)
(n=
24)
(n=
22)
(n=
41
)
AM
P$
(32)
2(8
)$
(36)
25(6
1)
CT
X0
(0)
0(0
)0
(0)
0(0
)
XL
N0(0
)0
(0)
0(0
)0(0
)
GE
N1
(4)
2(8
)1
(4)
12(2
9)
NE
O10
(40)
5(2
1)4
(18)
20(4
9)
KA
N10
(40)
6(2
5)$
(36)
19(4
6)
TE
T22
(88)
22(9
2)2
1(9
5)
39(9
5)
CH
L4
(16)
1(4
)2
(9)
17(4
1)
TM
P1(
4)2
(8)
2(9
)25
(61)
SU
L22
(88)
24(1
00)
22(1
00)
34(8
3)
Phe
noty
pic
resi
stan
cew
asov
eres
tim
ated
beca
use
only
isol
ates
show
ing
anti
mic
robi
alw
ere
furt
her
anal
yzed
.b
AT
M,
anti
mic
robi
als:
AM
P,
ampi
cill
in;
CT
X,
cefo
taxi
me;
XL
N,
ceft
iofi
ir;
GE
N,
gent
amic
in;
NE
O,
neom
ycin
;K
AN
,ka
nam
ycin
;C
HL
,ch
lora
mph
enic
ol;
TE
T,
tetr
acyc
line
;T
MP
,
trim
etho
prim
;SU
L,
sulf
onam
ides
.
DD
66
Tot
al
(n=
112)
43(3
8)
0(0
)
0(0
)
16(1
4)
39(3
5)
43(3
8)
104
(93)
24(2
1)
30(2
7)
102
(91)
init
ial
resi
stan
ceto
one
DD
D67
TA
BL
E3.
Dis
trib
uti
on
ofan
tim
icro
bia
lre
sist
ance
gene
sde
tect
edin
ET
EC
01
49
:K
91is
olat
esac
cord
ing
toth
eis
olat
ion
peri
ods
Num
ber
ofpo
siti
veis
olat
esac
cord
ing
toth
eis
olat
ion
peri
ods
(%)a
Ant
imic
robi
alA
ntim
icro
bial
Tot
alnu
mbe
r(¾
)of
197$
-198
419
85-1
989
1990
-199
419
95-2
000
resi
stan
cere
sist
ance
prob
epo
siti
veis
olat
es
Am
pici
llin
blaT
EM
8(1
00)
1(50
)3
(38)
25(1
00)
37(8
6)
blas
Hv
3(3
8)0
(0)
1(12
)5
(20)
9(21
)
bÏa
ox
0(0
)0
(0)
2(2
5)0
(0)
2(5
)
bÏao
xp7,
bÏap
sE,
0(0
)0
(0)
0(0
)0
(0)
0(0
)b1
aCT
XM
.3
Gen
tam
icin
aac(
3)-I
V0(
0)0
(0)
1(10
0)11
(92)
12(7
5)
ant(
2‘9
-la,
0(0)
0(0
)0(
0)0
(0)
0(0)
aac(
3)-l
Ia
Neo
myc
inap
h(3
9-fa
9(9
0)4
(80)
4(1
00)
17(8
5)34
(87)
aph
(39
-IIa
1(10
)1(
20)
2(50
)2(
10)
6(15
)
oc
- r—
_
oO N N
ON N
o r—- oo o
o
oo o
o
nH
O 00
O r—
o
ON N
oo
O
N —
oO
oo
_
OO r—O
o oO
O
O
O
o
oO
N
N
C)
Q-Qu
S’ S’N S
S’ ;‘ ‘-‘ S’N 5r—
ocl.)
o o o oo o o
N NCdDO.)
oS’
_
‘-‘S S S NO O O O
N N
-J
-J
cj
O O O O 00
IOs
-Co
oO)bf
n I J
DD
D70
TA
BL
E4.
Dis
trib
uti
on
ofcl
ass
1in
tegro
nam
ong
the
ET
EC
0149:
K91
isol
ates
Tot
alno
.%
ofin
tegr
oncl
ass
1-po
siti
veis
olat
esac
cord
ing
toth
eIs
olat
ion
No.
of
inte
gron
clas
s1-
ofle
ngth
ofth
eva
riab
lere
gion
peri
odpo
siti
veis
olat
esis
olat
es1.
0-kb
1.3-
kb1.
6-kb
1.8-
kb
1978
-198
425
1421
790
0
1985
-198
924
1429
720
0
1990
-199
422
1926
680
6
1995
-200
041
2035
1535
15
Tot
al11
267
2955
106
DD
D
TA
BL
E5.
Ass
ocia
tion
betw
een
the
varf
ous
anti
mic
robia
lre
sist
ance
gene
san
dcl
ass
1in
teg
ron
amon
gE
TE
C0149:
K91
isol
ates
.
71
Ant
imic
robi
alre
sist
ance
gene
san
dcl
ass
1in
tegr
ona
aph
(39
-ap
h(3
9bl
aTE
Mbl
asH
vaa
c(3)
-IV
dhif
ldh
ijV
d/!f
rXll
Isu
tisu
lIl
tet(
A)
tet(
3)
tet(
C)
catI
la-l
ia
blas
Hv
++
+
aph(
39-
la++
H-
-H
aph
(39
-Ila
--
-
aac(
39-
IV+
+±
++
+-
dhfr
l-
--
--
dhfr
V+
+±
+-
-+
++
-
dIf
rXII
I-
--
-±
--
situ
-—
--
--
--
sull
I+
--
--
--
-H-
-
tet(
A)
--
-+
--
+-
(H)b
-
tet(
B)
--
-(+
)-
--
-+
±+
-(+
+±
)
tet(
C)
--
-±
--
+-
(-H
-)-
+-H
-(+
+±
)
catI
±+
+-
--
+-
±+
--
-
Cla
ss
inte
gron
bii
yan
tim
icro
bial
resi
stan
cege
nes
that
exhi
bite
das
soci
atio
nw
ithan
othe
rge
neat
the
F<
0.05
leve
lar
esh
own.
Sig
nifi
canc
ele
vel
of
asso
ciat
ion
(as
asse
ssed
byth
ech
i-sq
uare
exac
tte
st):
-,F
>0.
05;
+,
0.05
?P
?0.
01;
-H-,
0.01
?:F
?0.
001;
++
+,
0.00
1?
P.
bP
aren
thes
esin
dica
teth
ene
gati
veas
soci
atio
ns.
Article II
Genetic analysis of antimicrobial resistance among extraintestinal
Escherichia cou isolates of animal and human origin
(Soumis à The Journal of Antimicrobial Chemotherapy)
73
Genetic analysis of antimicrobial resistance among extraintestinal Escherichia cou isolates of
animal and human origin
Christine Maynard’, Sadjia Bekal2, François Sanschagrin3, Roger C. Levesque3, Roland
Brousseau2, Luke Masson2, Serge Larivière1 and Josée Harell*
‘Département de Pathologie et Microbiologie, Faculté de Médecine Vétérinaire, Université de
Montréal, Saint-Hyacinthe, Québec, Canada, J2S 7C6, 2Biotechnology Research Institute,
Montréal, Québec, Canada, H4P 2R2, ‘ Pavillon C. - E. Marchand, Université LavaI, Québec,
Québec, Canada, G1K 7P4,
Key words: Antibioresistance, integron, urinary tract infection
*Coffesponding author. Mailing address: Département de Pathologie et Microbiologie,
Faculté de Médecine Vétérinaire, Université de Montréal, 3200, rue Sicotte, C. P. 5000, Saint
Hyacinthe, Québec, Canada, J2S 7C6. Phone: 450 773 $521 ext. $233. Fax 450 77$ $10$.
E-mail:
74
SYNOPSIS
Thirty-nine Escherichia cou strains collected in 2001 from different infected animal tissues,
16 human Escherichia cou strains isoiated from patient with urinary tract infections (UTI) in
1990 and 54 isolates from patient with UTI coiiected in 2001 from urine specimens were
characterised for their phenotypic and genotypic antimicrobial resistance profile. Among the
ten antirnicrobial agents tested, resistance was most frequent for ampiciliin, tetracycline and
suifonamides, whereas resistance to cephalosporins was present oniy among animal isolates.
Muitiresistant strains were observed among the animal and the human strains. The
distribution of the resistance genes for beta-lactams, aminoglycosides, chloramphenicol,
tetracycline, trimethoprim and sulfonamides was assessed by coiony hybridisation in ail
isolates. Significant differences in the distribution of tetracycline (tet(D)), chloramphenicol
(cati, catili, foR) and trimethoprim (dÏfrI, dhfrV, dIfr VII, difrXIIJ) resistance genes were
observed in the animal and human isolates. Difference between association of some resistance
and virulence genes in the animal and human isolates was demonstrated. The associations of
tet(A) with tet(C) and of suÏI with the class 1 integron were observed in both groups of
isolates. Thirty-three and thirty-one percent of the animal and human strains, respectively,
possessed a ciass 1 integron. four major different variable regions of class I integron were
ampiified which contained aminoglycosides (aadA], aadA2, aadA5, aadA6) and/or
trimethoprim (dhfrlb, dI?frXIL dIfrXVil) resistance genes, thus demonstrating the importance
of integrons in the emergence of resistance in E. cou isolated from animais and humans. In
conclusion, the animal and human ExPEC strains chosen in this smdy demonstrate distinct
resistance and virulence profiles. The animal strains seem to be a more heterogeneous group
than the human UTI strains.
75
INTRODUCTION
Escherichia cou is an important cause of animal and human disease worldwide 1 This
bacterium can be classified into 3 major groups: commensal strains, intestinal pathogenic
(enteric or diarrheagenic) strains, and extraintestinal pathogenic strains. Extraintestinal
infections (Els) due to E. coït, called extraintestinal pathogenic E. cou (ExPEC), are common
in animal and human and can involve almost any organ or anatomicai site 2 Typical Els
include urinary tract infection (UTI), meningitis, diverse intra-abdominal infections,
pneumonia, soft-tissue infection, intra-vascular-device infections and osteomyelitis 2
Antimicrobial therapy is an important tool to treat these infections. However, resistance to
existing antimicrobials is widespread and of concem to veterinary and human medicine
A close association exists between the usage of antimicrobial agents and the leveis of
resistance observed 6-8 The antimicrobial classes routinely used for treatment of human
infections are also used for animal either for therapy and prevention or as growth promotion
factor. food animal or pets provide favourable conditions for the spread in human of bacteria
such as E. coït , , 10 Because the same antimicrobials are used for humans, pets and food
animais, it is difficult to show the contribution of the resistant bacteria or genes from the
animais in the emergence of antimicrobial resistant E. cou in humans and vtce-versa 10
In cases where an antimicrobial agent is used for animais but not for humans, some
unes of evidence exist. For exampie, afier the introduction of streptothricin for growth
promotion in 1980s, E. coïi harbounng a piasmid mediating resistance appeared and were
isoiated from pigs, pig farmers and their families Another example is the emergence of
apramycin resistance after its introduction in veterinary medicine. Direct transfer of piasmids
mediating apramycin resistance from E. cou in pigs to E. coïi from humans has been observed12-14
Furthermore, the virulent strains of E. cou that cause extraintestinai infection in animal
and human have numerous phylogenetic, pathotypic and genotypic simiiarities 15, 16 Some
studies support the hypothesis that ExPEC represent overlapping populations, with members
of certain clones or clonai groups capable of causing infections in pets, food animals and
humans7, 15, 17, 18
The dissemination of antimicrobiai resistant and virulent strains among different hosts
are an increasing probiem in infectious diseases. Many antimicrobial resistance and virulence
genes are iocated on plasmids and on transposons, enabiing their transfer among a variety of
bacterial species. In recent years, a third mechanism ofresistance gene dissemination lias been
discovered. It involves a DNA element that mediates the integration of resistance genes by a
76
site-specific recombinational mechanism. This novel DNA element, now called an integron, is
found either as part of transposons of the Tn2] family or independently on several groups of
broad-host-range plasmids. Four different integron classes have been characterised and the
most frequent one among clinical isolates is the class 1 19, 20 Class integrons possess two
conserved segments separated by a variable region (VR), which includes integrated
antimicrobial resistance genes, or gene cassettes witli unknown functions 21 The 3’ conserved
segment contains the qacEAl and sut! genes and an open reading ftame (ORF), oij5. The
qacEz\1 and sulI genes determine resistance to ethidium bromide and quatemary ammonium
compounds and to sulfonamides, respectively 21
Before reviewing the use of antimicrobials in animal husbandry and their effect on
human health, it is helpful to determine the E. cou resistance gene profile of these two
population groups. The objective of this study is to compare the genotypes of acquired
ExPEC antimicrobial resistance. In this study, animal and human ExPEC collected from
animal tissues and from urine of human UTIs, were analysed for their resistance to
antimicrobials and the distribution of the related antimicrobial resistance genes. Also, the
strains were characterised for their ExPEC virulence gene profile and for their phylogenetic
group.
77
MATERIALS AND METHODS
Bacterial strains and growth conditions. In this study, only strains showing at least
one antimicrobial resistance were selected. 0f the Escherichia cou isolated in 2001 from
different infected animal tissues (lung, heart, liver, lddney, brain and other), only a small
number (39) showed resistance to at least one antimicrobial and were kept for this study.
Animal strains were isolated from swine (40%), chicken (20%), cattie (20%) and pets (20%).
Only 8 antimicrobial resistant E. cou strains isolated from urine specimens of pets could be
obtained for this study therefore to enlarge our study ExPEC from other animal origin were
included. Sixteen human E. cou strains isolated from patients with UTI in 1990 at Sainte
Justine Hospital in Montreal 22 and 54 isolates collected in 2001 at Honoré-Mercier Hospital
in Saint-Hyacinthe were analysed in this study. The reference strain, EschericÏzia cou ATCC®
25922, was used to test each lot of antimicrobials by disc diffusion method. The bacterial
strains, kept at -80°C in tryptic soy broth (TSB) medium containing 10% glycerol (v/v), were
cultured on tryptic soy agar (TSA) supplemented with 5% sheep blood (v/v).
The thirty-two strains used as positive controls and templates for DNA amplification
were obtained from different laboratories (Table I). These strains were kept at -80°C as frozen
stocks in tryptic soy broth (TSB) medium containing 10% glycerol (v/v) and were propagated
on Luria Bertani (LB) broth or agar containing one of these antimicrobials at the following
concentrations: ampicillin (50 tg/mL), gentamicin (30 jig/mL), kanamycin (50 .tg/mL),
tetracycline (10 j.tg/mL), chloramphenicol (10 .tg/mL), trimethoprim (10 j.tg/mL);
suÏfamethazine (200 tg/mL).
Antimicrobial susceptibility testing. Antimicrobial susceptibility testing was carried
out for all strains by disc diffusion method according to the recommendations reported by the
NCCLS 23 As recommended by the NCCLS, Mueller-Hinton (MH) agar was used as culture
medium. The antimicrobial agents used in this study were: beta-lactams: ampicillin (10 jig),
ceftiofur (30 jig), cefotaxime (30 tg); aminoglycosides: gentamicin (10 tg), kanamycin (30
g), neomycin (30 tg); tetracycline (30 jig); chloramphenicol (30 tg); trimethoprim (5 jig);
sulfaminoxazole (250 ig). The antimicrobial breakpoints were taken from the NCCLS
recommendations except for neomycin, for which the breakpoints used were those
recommended by the manufacturer.
PCR primers and amplification. Resistance gene primers were dcsigned using the
GCG software program “Prime” (Genetics Computer Group, Madison, WI, USA). Virulence
genes which are associated to the ExPEC strains, afa3, afaD8, sfaDE, papC, hÏyA and iucD,
were tested and their primer sequences were previously described 24-27 (Table I).
78
Oligonucleotide primers were synthesized with a model DNA Synthesizer (Biocorp Inc.,
Montréal, Québec, Canada). The class 1 integron is characterised by the qacEAl and sulI
genes at its 3’conserved segment 28 Primers located at the 3’ conserved segment were used as
described by Sandvang et al 29 to investigate the presence ofthe class 1 integron (Table I).
Amplification was performed on 5 tL of supematant from bacterial preparation boiled
for 10 minutes 22 The PCR reaction mix (total 50 tL) included 29.6 tL H20, 5 tL of 1OX
PCR buffer (Amersham Pharmacia Biotech Inc., Piscataway, NJ, USA), 5 jiL of dNTP 2 mM,
Ï U of Taq polymerase (Amersham Pharmacia Biotech Inc.), and 25 pmol of each primer.
DNA amplification was carried out in a GeneAmp® PCR system 9700 (Perkin-Elmer, Foster
City, CA, USA) by using the following conditions: 5 min at 94°C followed by 30 cycles of
94°C for 30 s, 50°C for 30 s and 72°C for 1.5 min. A sample of 3 jit of the PCR product
wasverified for size and purity by gel electrophoresis (1.2 % (w/v) agarose in iX TAE (Tris
acetate-EDTA) buffer).
AÏ! isolates that contained the 3’ conserved segment of the class 1 integron were
further investigated by another PCR amplification of a VR within the integron by using
primers described by Sandvang and Aarestrup 29(Table I). Amplification conditions for these
primers were as follows: 5 min at 94°C, followed by 35 cycles of 94°C for 1 mm, 50°C for 1
mi and 72°C for 5 min.
The phylogenetic group to which the ExPEC strains belong was determined by PCR
based method as described by Clermont et al 30 (Table I). The data of the three amplifications
resulted in an assignment of the strains to phylogenetic groups as follows: chuA-, TspE4.C2-,
group A; chuA-, TspE4.C2+, group BI; chuA+,yjaA+, group B2; chuA+,yjaA-, groupD 30
DNA sequencing. The amplificd products were purified with a QlAquick PCR
purification kit or QlAquick gel extraction purification kit (QIAgen Inc., Mississauga,
Ontario, Canada). Their sequences were confirmed with the dRhodamme Terminator Cycle
Sequencing Ready reaction Kit using a model 377 DNA sequencer (Applied Biosystems,
foster City, CA, USA). Sequences were submitted to the National Center for Biotechnology
Information (Bethesda, Md) for comparison with sequences in GenBank by use ofthe BLAST
program. Multiple DNA alignments were performed by using the CLUSTALW program 31
Colony hybridisation. The amplicons were labelled with [a-32P]CTP by using a DNA
labelling beads kit (Amersham Pharmacia Biotech Inc.). Colony hybridisations were
performed as described previously 32
Statistical methods. Comparisons of the associations between resistance genes, class
1 integron and virulence genes, in ExPEC animal sfrains and in ExPEC human strains
79
separately, were made by using Pearson’s chi-square exact test (SAS, version 8.2; SAS
Institute, Inc., CARY, N.C., USA). The statistical significance was set at a P value of <0.05.
An association between two genes can be positive indicating that the genes are found
together, or negative, indicating that the genes are flot found together.
80
RESULIS
Antimicrobial resistance phenotype characteristics. One hundred fine isolates (39from animal origin and 70 from human origin) have been selected and charactensed for theirantirnicrobial resistant phenotype to ten commonly used antimicrobials by disc diffusionmethod. The percentage of resistant E. cou strains to the different antimicrobial agentsisolated from tissues of animal and human UTI is presented in Table II. Some isolatesdemonstrated a zone of inhibition at the limit (the zone between resistance and susceptibility)for cephalosporins (ceftiofur and cefotaxime) and aminoglycosides. To facilitate our analysis,these isolates were considered as resistant in this study.
Similar pattems of resistance to the different antimicrobial agents between strainsisolated from animal or from human origin were observed for six out of the ten antimicrobialstested. The E. cou strains isolated from animais appeared more resistant to cephalosporins,neomycin, kanamycin and tetracycline. All animal strains resistant to ceftiofur were alsoresistant to cefotaxime. fifty-one percent of the animal isolates wcre resistant to at least fourantimicrobials and 43 of them had the ampicillin! tetracycline! trimethoprim/ stilfonamidesresistance profile (data flot shown). Only one pet strain was multiresistant. The majorityshowed only one resistant to ampicillin or tetracycline.
Globally, the majority of the human strains were resistant to ampicillin and none ofthestrains were resistant to cephalosporins. Human E. cou strains that were isolated in 1990 wereless resistant to antirnicrobials than those isolated in 2001. None of the strains from 1990 wereresistant to the aminoglycosides and chloramphenicol. Twenty-three percent of the humanisolates were resistant to the following four antimicrobials: ampicillin, tetracycline,trimethoprim and sulfonamides. In 2001, 57% of the resistant strains showed more than threeresistances to antimicrobials tested. Comparatively, in 1990, only 25% had more than threeresistances to antimicrobials tested (data flot shown).
Distribution of resistance genes. The choice of the studied resistance genes wasbased on their relative importance in resistant E. cou strains Therefore, 28 genesencoding for antimicrobial resistance belonging to six antimicrobial families (beta-lactams,aminoglycosides, tetracycline, phenicols, trimethoprim and sulfonamides) were chosen todetermine their distribution in E. cou strains isolated from extraintestinal infections fromanimal and human cases (Table I).
(j) Beta-lactams. The DNA hybridisation probes bÏaTEM and blasHv detect ail theknown variants within the corresponding blaTEM and bÏasHv gene families on colonyhybridisation. The probe blaox7 detects bÏaoxj variants such as blaoxAIo to btaoxAl4,
81
blaoxJ.16 to blaoxpi9, blaoxA2s, blaoxA31, blaox35 (80% to 96% of similarity) and bÏapsE2
(96.3% of similarity). To furtber discriminate variants among the b1aoXJ7 hybridisation
positive isolates, a PCR amplification was undertaken using the blaoxp7 specific primers. Ail
strains were positive for at least one of the tested genes except for four animal strains, five
human strains isolated in 1990 and two human strains isolated in 2001.
More beta-lactams resistant isolates obtained from humans were positive to the bÏa--EM
probe (88%) than animal isolates (68%) (Table III). Twelve percent of ampicillin resistant E.
cou strains isolated from humans possessed also a bÏasnv gene whereas none of the animal
isolates were bÏasHv-positive. Three animal strains possessed the blaoxpi gene and none of
the human strains possessed this gene. None of the strains were positive with blacTXM3,
blapsE4, and bÏaoxJ..7 probes. Eleven (eight animal isolates and three human isolates) out of
the 82 strains resistant to ampicillin did flot possess one of the tested beta-lactam genes. The
genes btaoxi and bÏaTEM were found among two cephalosporins resistant strains isolated
from anirnats (data not shown).
(ii) Aminoglycosides. 0f the five tested aminoglycoside resistance genes, ant(2 ‘9-la,
aac(3)-IV, and aph(3 9-lia were flot found among the resistant strains. The aph(3 9-la gene
resulting in a kanamycin and neomycin resistance phenotype, was twice as frequent in animal
isolates resistant to these two antimicrobials than the human isolates having the same
resistance pattems (Table III). A resistance gene, aac(3)-IIa, was found in only four out ofthe
15 gentamicin resistant strains.
(iii) Tetracyclines. 0f the six tested tetracycline resistance genes, tet(A), tet(B), tet(C)
and tet(D) were found among the isolates. The tet(A), tet(B) and tet(C) genes were found
among animal and human isolates in high frequency (Table III). The tet(A) and tet(C) genes
were found together in 75% ofisolates. The tet(D) gene was found only in two humans strains
isolated in 1990. Three animal and one human tetracycline resistant isolates were tet-negative.
(iv) Phenicols. 0f the four tested chloramphenicol resistance genes, catI, catiil, and
foR were found among the strains. Arnong the eight chloramphenicol resistant isolates found
in animais, the most frequent phenicol resistance gene wasftoR (63%) (Table III). Only one
strain possessed the catI gene and another catili gene. Among the 17 human isolates, the most
frequent chloramphenicol resistance gene was catI (82%). Only one human strain was foR-
positive (Table III). Two animal and three human chloramphenicol resistant isolates were
negative with the chloramphenicol probes. One out of the 31 chioramphenicol susceptible
animal isolates was foR-positive and one out of 55 chloramphenicol susceptible human
isolates was catI-positive.
$2
(y) Trimethoprim. Trimethoprim resistance phenotype was associated, in this study,to four genes: dhfrI dIfrV, dIfi-VII, and dhfrXIII (Table III). Whereas the gene dJJH iscommon in the different groups of isolates, dhfr VII was found exclusively in the 2001 humanisolates group and dÏfrV and dÏfrXIII more specifically in the isolates of animal origin.Among the animal resistant strains, three were negative for the tested genes and onesusceptible strain was positive for dhfrXIII. Among the 27 human resistant strains, seven werenegative for the tested trimethoprim resistance genes.
(vi) Sulfonamides. The suÏI and sutil genes were largely present among the animaland human isolates. Thirty-seven percent of sulfonamide resistant animal strains had both sutIand suÏII. Whereas 30% ofthe resistant human strains isolated in 2001 contained both suÏIand suÏIJ, the human strains isolated from 1990 contained either sutI or sulli. four out of the30 resistant animal strains were sul-negative, whereas two ofthe 22 susceptible human strainswere sulI-positive.
Identification and characterisation of integrons. The class 1 integron ischaracterised by the qacEA 1 and sulI genes at its 3’ conserved segment responsible for theresistance to disinfectants and sulfonamides respectively 28 These genes were used to identifyby PCR strains containing the 3’ consewed region. 0f the 109 isolates, 57 strains (52%)arnplified a PCR fragment from the amplification of the class 1 integron 3’ conserved region(Table IV). A similar percentage of class 1 integron positive strains was observed among theanimal and human isolates. Two of the class I integron positive animal strains were sulInegative.
The VR of class I integron positive strains was characterised using the primers Fintiand Rinti and among the 57 class 1 integron positive strains, five distinct amplicons of 0.5kb, 1.0 kb, 1.6 kb, 1.7 kb, and 1.8 kb were obtained. Two representatives of each VR length(except for 1.7 kb) were sequenced. Both VR of the same Iength contained the same genecassettes with more than 95% of similarity. The most frequent VR amplified fragments (i.e.1.0 kb) in animal and human strains (5$ and 59%, respectively) contained the ant(3 ‘)-Ia(aadA]) gene cassette (GenBank accession no. X12$70) encoding streptomycin andspectinomycin resistance. Ah human class 1 integron positive strains that were isolated in1990 had a 1.0 kb VR. In 21 % ofthe animal and 28% human strains, respectively, the size ofthe VR was 1.6 kb that contained two gene cassettes, dhfrlb, encoding for resistance totrimethoprim, and ant(3 “)-If(aadA6), encoding resistance to streptomycin and spectinomycin(GenBank accession nos. AF3935 10 and Af 140629, respectively). Only one strain isolatedfrom animal had a VR length of 1.7 kb. This VR possessed two gene cassettes named
83
dhfrXVII which confers a resistance to trimethoprim, and ant(3 ‘)-Ie (aadA5), which confers aresistance to streptomycin and spectinomycin (GenBank accession no. AF169041). Finally,the 1.8 kb VR fragments possessed two gene cassettes, dhfrXII and ant(3 ‘9-If, which conferresistance to trimethoprim and resistance to streptomycin and spectinomycin, and one offnamed orJF (GenBank accession nos. Z21672 and AF284063, respectively).
Distribution of virulence genes. The Table III shows the distribution of the ExPECvirutence genes, afa3, afaD8, sfaDE, papC, hlyA and iucD. More human strains were positivefor these genes than animal strains, with the exception of afaD8 and iucD that were twice asimportant in animal strains than human ones. The gene iucD was the most frequently detectedvirulence genes (77%) among the animal isolates. No specific virulence gene profile could beassigned for the animal isolates. In pet isolates sfaDf, papC, hÏyA and iucD were present in 6out of $ isolates (data flot shown). Only 8% of animal strains were negative for the testedvirulence genes and 23% had more than two ofthese genes.
The most frequent virulence gene found in human isolates (53%) was papC. Noimportant differences between the virulence profile of human strains isolated from 1990 and2001 were noticed. Only 13% of human isolates did flot have the tested virulence genes and54% had more than two ofthese genes.
Association between resistance and virulence genes. $tatistically significantassociations between the occurrence of individual resistance and virulence genes among thewhole collection of animal and human ExPEC isolates were detected (F< 0.05) (Table V andVI). In both groups of animal and human strains, the association of the gene sutI with thepresence of class 1 integron was clear as well as the association of tet(A) with tet(C) genes.Several different associations were evident among the animal and human ExPEC strains. Forexample, the positive association of blaTEM and dÏfrV and the negative association oftet(A)/tet(C) with tet(B) genes were observed only among the animal strains. Also, someassociations were observed only among human strains such as the positive association ofblaTEM with sulIl as well as sutI with tet(A)/tet(C) genes. In strains isolated from humans,hÏyA was associated with dIfrI and iucD with sulli. In animals strains, sfaDE was weaklyassociated with dhfrXfII, tetA and tetC. In both groups, hÏyA was associated with the presenceof papC but only with sfaDE in strains ftom humans. In strains isolated from humans, thegene iticD was positively associated with sfaDE and hÏyA while negatively associated withthese genes in strains from animals.
Phylogenetic groupïng of resuits. Phylogenetic analysis studies suggest that E. coucan be divided into of four main phylogenetic groups (A, Bi, B2 and D) and that more
84
virulent ExPEC strains mainly belong to groups B2 and D, whereas commensal and some
virulent strains belong to group A and Bi. Clermont et al 30 had described a simple and rapid
phylogenetic grouping technique based on PCR.
The phylogenctic analysis showed that 46% of the strains isoiated from animal tissues
belong to phylogenetic group A, 20% to D, 20% to 32 and 15% to Bi. Among the animal
isolates, the phylogenetic group A and BI were mostly (56% and 83% respectively) porcine
strains (data flot shown). More than 80% of these isolates were associated with sudden death
of piglets due either to primaiy or secondary septicaemia and from which E. cou were isolated
in significant numbers from intestine or extraintestinal organs . Most pet strains (88%)
belonged to phylogenetic group B2 and 63% of avian strains were in the group D.
The half of the strains (54%) of the human isolates belonged to phylogenetic group
B2, 23% to group D, 19% to group A and only 4% to group Bi. Human strains isolated from
1990 belonged to group B2 in majority (75%) and 25% belonged to group A. Interestingly,
most human strains isolated from 2001 belonged to either group B2 (45%) or D (45%) and a
small proportion to group A and Bi (7% and 3%), respectively. Most animal isolates (71%)
belonging to group B2 contained the virulence genes sfaDE, papC and hlyA; among human
isolates the proportion was only 39%. Human strains belonging to group A contained fewer
virulent genes than the other human isolates.
$5
DISCUSSION
E. cou causes diverse extraintestinal infections in animais and in humans 2 is the
most common cause of UTI, community-acquired bacteraemia and sepsis Treatment of
ExPEC infection commonly includes the use of broad-spectnim antimicrobials 40• In the
present smdy, ExPEC strains isolated from infected tissues of animal and from urine of
human UTI, were analysed for the distribution of related acquired antimicrobial resistance
genes and virulence genes, and were characterised for their class 1 integron profile and
phylogenetic group.
Our study dernonstrates that similar antibioresistance pattems could be observed in
animal and human isolates. Ampiciilin, tetracycline and sulfonamides were the antimicrobiai
agents to which most ExPEC strains were resistant. It is noteworthy that ampicillin and
sulfonarnides are old antimicrobiais, which are wideiy used. Considering that tetracycline is
much Iess used in humans than in animal, the relativeiy high resistance for tetracyciine among
the isolates of human origin is interesting (Table II). More animal strains were resistant to
kanamycin than human strains whiie this antimicrobiai agent is more widely used in human
therapy. This can be expiained by the fact that aminoglycoside resistance gene aph(3 ‘)-Ia,
found in majority among these strains, can hydrolyse both kanamycin and neomycin.
Therefore, the seiected pressure due by the use of neomycin in animais, particulariy in pigs,
would have concurrentiy selected neomycin and kanamycin resistant strains. The
muitiresistance to antimicrobials observed in haif of the isolates can be also expiained by the
co-selection of resistance genes by the use of some other antimicrobiais 41 For example, it
was shown that the use of injectable oxytetracyciine in caille receiving in-feed
chlortetracycline was associated with an increase in the prevaience of resistance to
chioramphenicoi and suifisoxazole 42•
E. cou strains from animais and humans with similar phenotypic pattems of
antimicrobial resistance had differences in their genotypic resistance gene pattems. For
example, whereas among the ampiciilin resistant strains, the oxa-] gene was found only in
animal strains, the blasnv-type determinant was found oniy in human strains. These two
determinants are rareiy found together among ampicillin resistant E. cou ‘. Ceftiofur is a
cephalosporin used in food-producing animals in Canada and in the United States.
Cephalosporin resistant isoiates were found among animai origin strains but not from human
origin. Our smdy seems to indicate that resistance to cefiiofur was associated with other beta
iactam genes than those tested.. The cephamycinase blacMy genes that were not inciuded in
our study were found in E. cou showing a cefiiofur-cefiriazone resistance phenotype In the
86
case of tetracycline resistant strains, the tet(D) gene was exclusively associated with the
human strains origin. furthermore, the tet(D) gene was only found in strains isolated in 1990.
A more striking observation was that the aminoglycosides resistance gene aph(3 )-Ia was
present in kanamycin resistant human isolates twice less frequently than in animal isolates. It
appeared that kanamycin resistance in human strains was due in majority to other resistance
genes not tested in this study such as aac(3)JII or aph(3 )- VI that were observed in other
studies46. Also, the majority to the chloramphenicol resistant human strains were due to the
catI gene, whule the majority of chloramphenicol resistant animal strains was associated with
foR. In spite of a ban on the use of chloramphenicoi in food animais in Canada since 1980
chloramphenicol resistance in E. cou lias been also observed in various animal species by
others , In these studies, resistance to choramphenicol was closely associated with the
presence ofthe catI gene. In another study done by Bischoff et al [Bischoff, 2002 #202], only
four out of 4$ chloramphenicol-resistant isolates were harbouring a cadI gene; the relatively
unknown gene, cmlA, was responsible for the resistancc of the other isotates. The foR gene is
responsible for a cross-resistance between chloramphenicol and florfenicol, the latter although
used in animal therapy is not used in human medicine 6 Other studies report the presence of
foR in a large number of E. cou isolated from chickens and cattie 38, 5], 52 finally, the gene
dhfr VII which is the most frequent trimethoprim resistant gene found in the strains isolated
from humans was absent in strains from animais; on the other hand, the d1frV and dhfrXIII
the most frequent trimethoprim resistance genes in strains from animals were found in oniy
one strain from humans.
Nevertheless, some resistance genes were equally represented in animai and human
strains. It is shown by the Table III that the most frequent mechanism implicated of beta
lactams resistance in ExPEC from animal and human isolates was blaTEM-type beta-lactamase.
This is also documented by other studies &, “. These studies showed that mutations at the
level of the arnpC operon are another important cause of ampiciliin resistance in E. cou. This
couid be the case for eight beta-lactam resistant strains that did not possess any of the
acquired beta-lactam resistance genes tested. The resistance to sulfonamides in this study was
associated to two resistance genes, suil and sulil which animal and human sulfonamide
resistant strains possessed in similar proportion.
Resistance genes can be associated with mobile DNA like plasmids, transposons and
integrons known to facilitate their distribution . Half of the tested strains (52%) possessed
a ciass 1 integron. Because the integrons can inciude many different gene cassettes between
insertion sites at the VR, site-specific insertion of gene cassettes represents another further
87
mechanism that contributes to the evolution of plasmids and transposons of gram-negative
bacteria. Similarly to another study done by Bass et al , the most frequent amplified VR
length among the integron -positive avian pathogenic E. cou strains was 1.0 kb. In our study,
most class 1 integron in our E. cou collection, which possessed the same VR lengtli, appeared
to have acquired the same gene cassettes into their integrons. Although multiple gene cassette
insertions and over 40 distinct gene cassettes have been found linked to integrons 21 only five
different gene cassette sizes were detected among E. cou isolates in this study, and have been
also reported in other studies 56, Furthermore, except for the sull gene, gene cassettes found
among the sequenced class 1 integron did not correspond to the tested resistance genes in this
study. The presence of integrons emphasizes their potential to contribute to efficient spread of
antibiotic resistance. One clear observation arising from our study is that the number and
diversity of genes driving the phenotypic resistance are dynamic. Some genes associations
were observed. Sirnilarly to other studies, the sulI gene was associated with the class 1
integron 28, 58 As in our previous study on ETEC strains isoiated in the Province of Quebec
from pigs (Maynard, C., et al., 2003 In press.), the association between tet(A) and tet(C) has
also been observed. The blaTEM genes isolated from animal strains origin were associated with
dhfrV, whereas in strains isolated from humans, they were associated with the sulIl gene.
Associations between genes could help the understanding of the certain genes presence in the
light of co-selection or transmission. for example, because the association of blaTEM genes
with either dhfrV or sulil depending on the origin of the isolates could suggest that strains
from animal and human origin acquired different genes. Aithough some studies suggested
that, in addition to the gene transfer between two different bacterium, bacteria couid transfer
its genetic material to another bacteria from to different hosts I 1-14 In these cited studies, the
acquired antimicrobial resistance genes were found on an integron or a plasmid that showing
that the mobile DNA element importance in these resistant deterrninant spread. ThefloR gene
presence in strains isolated from animais and in a strain isolated from human can indicate
either gene transmission or resistant strain transfer between the two hosts. It would be
tempting to suggest that the foR gene observed in one strain isolated from human is
indicative of a gene transfer or of a resistant strain transfer from animais to human. The foR
gene was reported in SalrnoneÏla isolates and in an old Kiebsietia strain
Although the virulence gene characterisation was not exhaustive, most E. cou isolates
in this study possessed some traits of extraintestinal strains. The relative virulence marker
representation is different in E. cou isolated from animals and humans. Among ExPEC
strains, pap (pyelonephritis-associated pili), sfii (S fimbrial adhesin) and afa (afimbrial
$8
adhesin) operons, encoding P, S, and Afa adhesins, respectively, are most commoniy found 60
In our study, the pap sequences were found in haif of animal strains and in of human strain,
and of these many, 22% were positive to more than one adhesin. More human strains
possessed hÏyA than animal strains, and among the latter, hlyA was mostly urine isolates from
pet animais. Other studies showed the important ftequency of hlyA among ExPEC isolated
from humans and animais. The aerobactin was frequently found in our collection.
Interestingly, the mai ority of animai E. cou isoiated from tissues were iucD (93%)
comparatively to animai strains from urine isolates (7%). $imilar resuits were reported by
Opal et ai 61 among human isolates . Virulence genes are also associated with mobile
elements. In our study, associations between the virulence genes hlyA and iucD and the
resistance antimicrobial genes such as sulil and dÏfrI were noticed in human strains. The
aerobactin genes are associated with ISJ-like elements and are found on large plasmids, often
CoIV-like, or chromosome in ExPEC 62 Similarly the hly determinants can be associated to
either piasmids or chromosome of E. cou 63 Strong associations between hly, papC and
sfaDE genes as well as with iucD and sfaDE genes were observed in UTI strains of human
ongin. Sirnilar associations were also observed in other studies on ExPEC 15, 64 Virulence
genes are found in pathogenicity islands (PAIs) consisting of particular genomic regions
carrying-associated genes together with the loci, the mobility of some PAIs has been
demonstrated 65 Moreover, different UTI strains such as 536, CFTO73, J96 contain PAIs
carrying both hly and P determinants.
Phylogenetic analysis described by Herzer et al 66 have shown that virulent
extraintestinal strains belong mainly to group B2 and, to a lesser extent, to group D 67, 68
whereas most commensal strains but also some virulent strains belong to group A and Bi.
While the majority of the UTI human isolates of our study belonged to virulent groups (B2 or
D) as found in other studies 15, 30 But it is flot the case of the strains isolated from animal
tissues. Interestingly, the majority of the pet strains belonged to phylogenetic group B2.
Aithough they share common traits, in general the animal and human ExPEC strains in this
study demonstrate a distinct resistance and virulence gene pattem. Although our ExPEC
collection originating from different host and tissues share some properties with strains from
animais are more heterogeneous than human UTI strains. Consequently, it is difficult to
establish links between the presence of given genes in both population groups . Interestingly,
TiTI strains from pets showed virulence and resistance gene profile similar to human TiTI
strains and most, in the two cases, belong to the pathogenic group B2. This commonality
89
between pets and human ExPEC was demonstrated by Johuson et aÏ 15, 64 The authors
suggested that infectious transfer between hosts of different origin can imply.
Our genotypic resistance analyses of extraintestinal isolates from different hosts show
that the genes behind phenotypic resistance are flot static but rather in a state of flux driven by
various selection forces such as the use of specific antimicrobials. Often, more than one gene
was associated with a given phenotypic resistance. This study reinforces the necessity of
using genotypic resistance analyses in future epidemiological studies.
90
ACKNOWLEDGMENTS
We are very gratefiil to Guy Beaucharnp for statistical analysis, to Dr. Madeleine
Fortin and Saint-Hyacinthe Hospital for their implication in of the strains collect and to Kim
Messier for phenotypic resistance characterisation of strains. We appreciate the teclmical
assistance provided by the members of the Groupe de Recherche sur les Maladies du Porc
(GREMIP).
This work was supported by the Natural Sciences and Engineering Research Council
of Canada (NSERC) under research work unit of Canadian Research Network on Bactenal
Pathogens of Swine grant 225 155 and by the Fédération des Producteurs de Porcs du Québec
(FPPQ).
91
REFERENCES
1. Salyers, A. A. & Whitt, D. D. (1994). Escherichia cou gastro-intestinal infections. In:
Bacterial pathogenesis: a inolecular approach. (Salyers, A. A. & Whitt, D. D., Eds),
pp. 190-204. American Society of Microbiology Press, Washington, D. C.
2. Russo, T. A. & Joimson, J. R. (2000). ProposaI for a new inclusive designation for
extraintestinal pathogenic isolates of Escherichia cou: ExPEC. The Journal of
i,fectious diseases 181, 1753-1754.
3. Aubry-Damon, H. & Courvalin, P. (1999). Bacterial resistance to antimicrobial agents:
selected problems in France, 1996 to 1998. Ernerging infectious diseases 5, 315-320.
4. Monroe, S. & Polk, R. (2000). Antimicrobial use and bacterial resistance. Current
opinion in microbioÏogy 3,496-501.
5. Teshager, T., Herrero, I. A., Ponero, M. C., Garde, J., Moreno, M. A. & Dominguez,
L. (2000). Surveillance of antimicrobial resistance in Escherichia coli strains isolated
from pigs at Spanish slaughterhouses. International journal of antimicrobial agents
15, 137-142.
6. Schwarz, S. & Chaslus-Dancla, E. (2001). Use of antimicrobials in veterinary
medicine and mechanisms ofresistance. Veterinay research 32, 20 1-225.
7. Aarestrup, F. M. & Wegener, H. C. (1999). The effects of antibiotic usage in food
animais on the development of antimicrobiai resistance of importance for humans in
Cainpylobacter and Escherichia cou. Microbes and i,fectio,i /Institut Pasteur 1, 639-
644.
8. Guillemot, D. (1999). Antibiotic use in humans and bacterial resistance. Current
opinion in rnicrobiology 2, 494-498.
9. van den Bogaard, A. E. & Stobberingh, E. E. (2000). Epidemiology of resistance to
antibiotics. Links between animais and humans. Internationaljournal ofantimicrobiat
agents 14, 327-335.
10. Heaith Canada (2002). Uses of antimicrobial drugs in food animais. In: Used of
antimicrobials in food animals in Canada : Impact on resistance and human health),
pp. 53-67. Heaith Products and food branch, Canada.
11. Hummel, R., Tschape, H. & Witte, W. (1986). Spread of piasmid-mediated
nourseothricin resistance due to antibiotic use in animai husbandry. Journal of basic
inicrobiology 26, 461-466.
92
12. Hunter, J. E., Bennett, M., Hart, C. A., Shelley, J. C. & Walton, J. R. (1994).
Aprarnycin-resistant Escherichia cou isolated from pigs and a stockman.
EpidemioÏogy and infection 112, 473-480.
13. Hunter, J. E., Hart, C. A., Shelley, J. C., Walton, J. R. & Bennett, M. (1993). Human
isolates of apramycin-resistant Escherichia cou which contain the genes for the
AAC(3)IV enzyme. Epidemiotogy and iifection 110, 253-259.
14. Jolmson, A. P., Bums, L., Woodford, N., Threlfall, E. J., Naidoo, J., Cooke, E. M., et
al. (1994). Gentarnicin resistance in clinical isolates of Escherichia cou encoded by
genes ofveterinary origin. Journal ofmedicat microbioÏogy 40, 22 1-226.
15. Jolmson, J. R., Steil, A. L., Delavari, P., Murray, A. C., Kuskowski, M. & Gaastra, W.
(2001). Phylogenetic and pathotypic similarities between Escherichia cou isolates
from urinai’,’ tract infections in dogs and extraintestinal infections in humans. The
Journal ofiïfectious diseases 183, 897-906.
16. Feria, C., Machado, J., Duarte Coneia, J., Goncalves, J. & Gaastra, W. (2001).
Distribution of papG alleles among uropathogenic Escherichia cou isolated from
different species. FEMS rnicrobiology letters 202, 205-208.
17. van den Bogaard, A. E., London, N., Driessen, C. & $tobberingh, E. E. (2001).
Antibiotic resistance of faecal Escherichia coli in poultry, poultry farmers and poultry
slaughterers. The Journal ofantimicrobial chemotherapy 47, 763-771.
1$. Low, D. A., Braaten, B. A., Ling, G. V., Johnson, D. L. & Ruby, A. L. (198$).
Isolation and comparison of Escherichia cou strains from canine and human patients
with urinary tract infections. Infection and irnrnunity 56, 260 1-2609.
19. Sallen, B., Rajoharison, A., Desvarenne, S. & Mabilat, C. (1995). Molecular
epidemiology of integron-associated antibiotic resistance genes in clinical isolates of
enterobacteriaceae. Microbial drug resistance (Larchmont, N. Y) 1, 195-202.
20. Jones, M. E., Peters, E., Weersink, A. M., Fluit, A. & Verhoef, J. (1997). Widespread
occurrence of integrons causing multiple antibiotic resistance in bacteria. Lancet 349,
1742-1743.
21. Recchia, G. D. & Hall, R. M. (1997). Origins of the mobile gene cassettes found in
integrons. Trends Microbiol 5, 3 89-394.
22. Daigle, F., Harel, J., Fairbrother, J. M. & Lebel, P. (1994). Expression and detection of
pap-, sfa-, and afa-encoded fimbrial adhesin systems among uropathogenic
Escherichia cou. Canadian jottrnal ofrnicrobiology 40, 286-291.
93
23. Watts, J. L., Chengappa, M. M., Cole, J. R, Cooper, J. M., Inzana, T. J., Plaunt, M. R.,
et al. (1999). Peiformance standards for antimicrobial disk susceptibiÏity tests for
bacteria isoÏated from animals, Approved standard document, M3 1-A edn, National
Committee for Clinical Laboratory Standards, Wayne, Pennsylvania.
24. Le Bouguenec, C., Archambaud, M. & Labigne, A. (1992). Rapid and specific
detection of the pap, afa, and sfa adhesin-encoding operons in uropathogenic
Escherichia cou strains by polymerase chain reaction. Journal of clinical
microbioÏogy 30, 1189-1193.
25. Lalioui, L., Jouve, M., Gounon, P. & Le Bouguenec, C. (1999). Molecular cloning and
characterization of the afa-7 and afa-8 gene clusters encoding afimbrial adhesins in
Escherichia cou strains associated with dianhea or septicemia in calves. Infection and
iinmunity 67, 5048-5059.
26. Yamamoto, S., Terai, A., Yuri, K., Kurazono, H., Takeda, Y. & Yoshida, 0. (1995).
Detection of urovimlence factors in Eschericïtia cou by multiplex polymerase chain
reaction. FEMS immunoÏogy and medical ,nicrobioÏogy 12, 85-90.
27. Henero, M., de Lorenzo, V. & Neilands, J. B. (1988). Nucleotide sequence of the
iucD gene of the pColV-K30 aerobactin operon and topology of its product studied
with phoA and tacZ gene fusions. Journal ofbacteriology 170, 56-64.
28. Paulsen, I. T., Littlejolm, T. G., Radstrom, P., Sundstrom, L., Skold, O., Swedberg, G.,
et al. (1993). The 3’ conserved segment of integrons contains a gene associated with
muttidrug resistance to antiseptics and disinfectants. Antimicrobial agents and
chemotherapy 37, 76 1-768.
29. Sandvang, D. & Aarestrup, F. M. (2000). Characterization of aminoglycoside
resistance genes and class 1 integrons in porcine and bovine gentamicin-resistant
Escherichia coli. Microbial drug resistance (Larchmont, N. Y) 6, 19-27.
30. Clermont, O., Bonacorsi, S. & Bingen, E. (2000). Rapid and simple determination of
the Escherichia coli phylogenetic group. AppÏied and environmental microbioÏogy 66,
4555-4558.
31. Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improvingthe
sensitivity of progressive multiple sequence alignment through sequence weighting,
position-specific gap penalties and weight matrix choice. Nucleic acids research 22,
4673-4680.
32. Harel, J., Lapointe, H., Fallara, A., Lortie, L. A., Bigras-Poulin, M., Lariviere, S., et aï.
(1991). Detection of genes for fimbrial antigens and enterotoxins associated with
94
Escherichia coli serogroups isolated from pigs with diarrhea. Journal of clin ical
microbiology 29, 745-752.
33. Bush, K., Jacoby, G. A. & Medeiros, A. A. (1995). A functional classification scheme
for beta-lactamases and its correlation with molecular structure. Antirnicrobial agents
and chemotherapy 39, 121 1-1233.
34. Chopra, I. & Roberts, M. (2001). Tetracycline antibiotics: mode of action,
applications, molecular biology, and epidemiology of bacterial resistance.
MicrobioÏogy and inotecutar biotogy reviews MMBR 65, 232-260 ; second page,
table of contents.
35. Huovinen, P., Sundstrom, L., Swedberg, G. & Skold, 0. (1995). Trimethoprim and
sulfonamide resistance. Antimicrobial agents and chemotherapy 39, 279-289.
36. Muer, G. H., Sabatelli, F. J., Hare, R. S., Glupczynsld, Y., Mackey, P., Shlaes, D., et
al. (1997). The most frequent aminoglycoside resistance mechanisms--changes with
time and geographic area: a reflection of aminoglycoside usage pattems?
Aminoglycoside Resistance Study Groups. Clinical i,fectious diseases . an official
publication ofthe Infectious Diseases $ociety ofAmerica 24 Suppl 1, S46-62.
37. Murray, I. A. & Shaw, W. V. (1997). O-Acetyltransferases for chloramphenicol and
other natural products. Antimicrobial agents and chernotherapy 41, 1-6.
38. White, D. G., Hudson, C., Maurer, J. J., Ayers, S., Zhao, S., Lee, M. D., et al. (2000).
Characterization of chloramphenicol and florfenicol resistance in Escherichia coli
associated with bovine dianhea. Journal ofclinical microbiology 38, 4593-4598.
39. Johnson, J. R. & Russo, T. A. (2002). Extraintestinal pathogenic Escherichia cou: “the
other bad E cou”. The Journal oflaboratoiy and clinical medicine 139, 155-162.
40. Bertchinger, H. U. & Fairbrother, J. M. (1999). Escherichia cou Infections. In:
Diseases ofswine, $ edn. (Straw, B. E., D’Allaire, S., Mengeling, W. L. & Taylor, D.
J., Eds), pp. 43 1-468. Iowa State University Press, Ames.
41. Alekshun, M. N. & Levy, S. B. (2000). Bacterial drug resistance: Response to survival
threats. In: Bacteriat stress responses. (Storz, G. & Hengge-Aronis, R., Eds), pp. 323-
366. ASM Press, Washinton, D. C.
42. O’Connor, A. M., Poppe, C. & McEwen, S. A. (2002). Changes in the prevalence of
resistant Escherichia cou in cattie receiving subcutaneously injectable oxytetracycline
in addition to in- feed chlortetracycline compared with caille receiving only in-feed
chlortetracycline. Canadian journal of veterinaîy research - Revue canadienne de
recherche veterinaire. 66, 145-150.
95
43. feria, C., Ferreira, E., Correia, J. D., Goncalves, J. & Canica, M. (2002). Pattems and
mechanisms of resistance to beta-lactams and beta- lactamase inhibitors in
uropathogenic Escherichia cou isolated from dogs in Portugal. The Journal of
antimicrobial chemotherapy 49, 77-85.
44. Brinas, L., Zarazaga, M., Saenz, Y., Ruiz-Larrea, F. & Torres, C. (2002). beta
Lactamases in Ampicillin-Resistant EschericÏiia cou Isolates from Foods, Hurnans,
and Heaithy Animais. Antimicrobial agents and chemotherapy 46, 3156-3163.
45. Zhao, S., White, D. G., McDermott, P. F., Friedman, S., English, L., Ayers, S., et al.
(2001). Identification and expression of cephamycinase bta(CMY) genes in
Escherichia coli and Salmonella isolates from food animais and ground meat.
Antimicrobial agents and chemotherapy 45, 3647-3650.
46. Shaw, K. J., Rather, P. N., Hare, R. S. & Miller, G. H. (1993). Moiecuiar genetics of
aminogiycoside resistance genes and familial reiationships of the aminogiycoside
modifying enzymes. Microbiological reviews 57, 138-163.
47. Giimore, A. (1986). Chloramphenicoi and the poiitics of heaith. CMAJ: Canadian
Medical Association jottrnal - Journal de l’Association inedicale canadienne 134, 423,
426-428, 43 3-425.
48. Bischoff, K. M., White, D. G., McDermott, P. f., Zhao, S., Gaines, S., Maurer, J. J., et
al. (2002). Characterization of chloramphenicol resistance in beta-hemolytic
Escherichia cou associated with diarrhea in neonatal swine. Journal of clin ical
microbiology 40, 3 89-394.
49. Lanz, R., Kuhnert, P. & Boeriin, P. (2003). Antimicrobiai resistance and resistance
gene determinants in ciinical Escherichia coli from different animal species in
Switzeriand. Veterinary rnicrobiology 91, 73-84.
50. Werckenthin, C., Seidi, S., Riedi, J., Kiossis, E., Wolf, G., Stolla, R., et al. (2002).
Escherichia coli isolates from young calves in Bavaria: in vitro susceptibilities to 14
anti-microbial agents. Journal of veterinaly medicine. B, Infectious diseases and
veterinaiy public health. 49, 61-65.
51. Keyes, K., Hudson, C., Maurer, J. J., Thayer, S., White, D. G. & Lee, M. D. (2000).
Detection of florfenicol resistance genes in Escherichia cou isoiated from sick
chickens. Antimicrobial agents and chemotherapy 44, 42 1-424.
52. Cloeckaert, A., Baucheron, S., Flaujac, G., Schwarz, S., Kehrenberg, C., Martel, J. L.,
et al. (2000). Piasmid-mediated florfenicoi resistance encoded by the foR gene in
96
Escherichia cou isolated from cattie. Antimicrobiat agents and chemotherapy 44,
2858-2860.
53. Jacoby, G. A. (1994). Extrachromosomal resistance in gram-negative organisms: the
evolution ofbeta-lactamase. Trends MicrobioÏ 2, 357-360.
54. Tenover, F. C. & Rasheed, J. K. (1998). Genetic methods for detecting antimicrobial
and antiviral resistance genes. In: Manuat of clinical rnicrobiology, 7 edn. (Murray, P.
R., Ed), pp. 157$-1592, Washington, D.C.
55. Bass, L., Liebert, C. A., Lee, M. D., Summers, A. O., White, D. G., Thayer, S. G., et
al. (1999). Incidence and characterization of integrons, genetic elements mediating
multiple-drug resistance, in avian Escherichia cou. Antirnicrobial agents and
chernotherapy 43, 2925-2929.
56. Roy, P. H. (1997). Dissémination de la résistance aux antibiotiques : le génie
génétique à l’oeuvre chez les bactéries. Médecine/$ciences 13, 927-933.
57. Schmitz, F. J., Hafner, D., Geisel, R, Follmann, P., Kirschke, C., Verhoef, J., et al.
(2001). Increased prevalence of class I integrons in Escherichia cou, KÏebsiella
species, and Enterobacter species isolates over a 7-year period in a German university
hospital. Journal ofcÏinical microbiology 39, 3724-3726.
5$. Levesque, C., Piche, L., Larose, C. & Roy, P. H. (1995). PCR mapping of integrons
reveals several novel combinations of resistance genes. Antimicrobial agents and
chemotherapy 39, 185-191.
59. Meunier, D., Baucheron, S., Chaslus-Dancla, E., Martel, J. L. & Cloeckaert, A.
(2003). florfenicol resistance in Salmonella enterica serovar Ncwport mediated by a
plasmid related to R55 from Kiebsiella pneurnoniae. The Journal of antimicrobial
chernotherapy 51, 1007-1009.
60. Smyth, C. J., Marron, M. & Smith, S. G. J. (1994). Fimbriae of Escherichia cou. In:
Escherichia cou in domestic anirnals and hurnans. (Gyles, C. L., Ed), pp. 399-435.
CAB international, Guelph.
61. Opal, S. M., Cross, A. S., Gemski, P. & Lyhte, L. W. (1990). Aerobactin and alpha
hemolysin as virulence determinants in Escherichia coli isolated from human blood,
urine, and stool. The Journal of infectious diseases 161, 794-796.
62. Bindercif, A. & Neilands, J. B. (1985). Aerobactin genes in clinical isolates of
Escherichia cou. Journal ofbacteriology 161, 727-735.
97
63. Donnenberg, M. S. & Welch, R. A. (1996). Virulence determinants of uropathogenic
Escherichia cou. In: Urinaîy tract iîfections, H. L. T. Mobley and J. W. Warren ASM
press edn), pp. 135-174.
64. Johnson, J. R., Delavari, P., Kuskowski, M. & Steil, A. L. (2001). Phylogenetic
distribution of extraintestinal virulence-associated traits in Escherichia cou. The
Journal of infectious diseases 183, 78-88.
65. Hacker, J. & Kaper, J. 3. (2000). Pathogenicity islands and the evolution of microbes.
Annual review of microbioÏogy 54, 64 1-679.
66. Herzer, P. J., Inouye, S., Inouye, M. & Whittam, T. 5. (1990). Phylogenetic
distribution of branched RNA-linked multicopy single- stranded DNA among namral
isolates of Escherichia cou. Journal ofbacteriology 172, 6175-6181.
67. Boyd, E. f. & Harti, D. L. (199$). Chromosomal regions specific to pathogenic
isolates of Escherichia coli have a phylogenetically clustered distribution. Journal of
bacteriology 180, 1159-1165.
68. Bingen, E., Picard, B., Brahimi, N., Mathy, S., Desjardins, P., Elion, J., et al. (1998).
Phylogenetic analysis of Escherichia cou strains causing neonatal meningitis suggcsts
horizontal gene transfer ftom a predominant pool of highly virulent 32 group strains.
The Journal ofinfectious diseases 177, 642-650.
DD
D
TA
BL
EI.
PC
Rpri
mer
sus
edfo
ran
tim
icro
bia
lre
sist
ance
gene
ampl
ific
atio
ns
98
b/aT
EM
btas
Hv
btao
xp,
bla
ox.7
blap
sE4
bla
cT
x3
ant(
2‘9
-la
(aad
B)”
aac(
3)-l
ia
(aac
C2)
aac(
3)-I
V
aph(
39-
la
fap
hA1)
aph’
39
-lia
(ap
hA2)
tet(
A)
let(
B)
tet(
C)
tet(
D)
tet(
E)
tet(
Y)
catI
GA
GT
AT
TC
AA
CA
TT
fTC
GT
TC
GC
CT
GT
GT
AT
FA
TC
TC
CC
GC
AG
CG
CC
AG
TG
CA
TC
AA
C
AG
TT
CT
CT
GC
CG
AA
GC
C
CT
GC
TC
GT
AT
AG
GT
GT
fl’C
C
AA
TC
AC
TG
CG
TC
AG
TT
CA
C
TC
CA
GA
AC
CT
rGA
CC
GA
AC
CG
GA
AG
GC
AA
TA
AC
GG
AG
GT
GIG
CT
GC
TG
GT
CC
AC
AG
C
AT
GG
GC
TC
GC
GA
TA
AT
GT
C
GA
AC
AA
GA
TG
GA
TT
GC
AC
GC
GT
GA
AA
CC
CA
AC
AT
AC
CC
C
CC
TF
AT
CA
TG
CC
AG
TC
TÏG
C
AC
HG
GA
GC
CA
CT
AT
CG
AC
TG
GG
CA
GA
TG
GT
CA
GA
TA
AG
flA
AT
GG
CA
AC
AG
CC
AG
C
AC
CG
CA
CT
CA
flG
TF
GT
C
AG
TT
GC
TC
AA
TG
TA
CC
TA
TA
AC
C
AC
CA
AT
GC
TF
AA
TC
AG
TG
A
CG
CA
GA
TA
AA
TC
AC
CA
CA
AT
G
CC
GC
AT
CA
AA
TG
CC
AIA
AG
TG
TC
TC
AA
CC
CA
AC
CA
AC
CC
TC
GC
AT
CM
TfC
GC
TC
TC
TI7
AT
CC
CC
CA
CA
AC
CC
AG
GC
AA
GA
CC
TC
AA
CC
ITF
tCC
GC
TC
TT
CA
GC
AA
TA
TC
AC
GG
GA
AG
GC
AA
GC
AG
GA
TG
TA
G
AC
TG
CC
Gli
nI
ICG
CC
CT
AC
AA
TC
CA
TG
CC
AA
CC
C
CA
GC
AC
AC
CC
TG
TA
G1T
TT
C
TC
CA
TA
CC
CA
TC
CA
flC
CA
C
TF
CC
AA
GC
AG
CA
AC
AC
AC
TF
GT
AA
TT
CA
flA
AG
CA
TT
CT
GC
C
857
AF
3098
24
768
Af1
4885
0
198
AJ2
3834
9
591
X75
562
705
J051
62
701
X92
506
R.
C.
Lev
esqu
e
R.
C.
Lev
esqu
e
Pas
teur
Inst
itute
R.
C.
Lev
esqu
e
R.
C.
Lev
esqu
e
A.
Hul
etsk
y
Ant
imic
robi
alG
enet
ic.
Am
plic
onG
enba
nkfo
rwar
dP
CR
prim
erse
quen
ce(S
’-3’
)R
ever
seP
CR
prim
erse
quen
ce(S
’-3’
)So
urce
ofD
NA
fam
ily
mar
kers
size
(bp)
acce
ssio
nno
.
Bet
a-L
acta
ms
Am
inog
lyco
side
s
Tet
racy
clin
e
Phe
nico
ls
TC
GA
AC
AG
GT
AG
CA
CT
GA
G
AG
[TG
AC
CC
AG
GG
CT
GIC
GC
CT
CA
CC
GA
GG
CA
GT
TC
CA
T
700
X04
555
R.
C.
Lev
esqu
e
740
X54
723
D.
San
dvan
g
627
X01
385
J.H
arel
600
M18
329
J.H
arel
680
V00
618
J.H
arel
888
X00
006
J.H
arel
774
J018
30J.
Har
el
881
J017
49J.
Har
el
827
X65
876
S.B.
Lev
y
853
L06
940
M.
C.
Rob
erts
823
AF
0709
99M
.C
.R
ober
ts
547
M62
822
J.H
arel
cati
l
cati
il
foR
dhfr
l
dhfr
V
dhfr
VII
dhfr
lX
dhfr
XII
I
sulI
sulI
l
afa3
afaD
8
sfaD
E
papC
htyA
iucD
AC
AC
GC
CC
AT
CG
TC
TT
CG
CC
GT
GA
GC
AIT
FT
G
CG
CC
GT
CA
TF
CC
TC
AC
CT
TC
AA
GA
AT
GG
AG
TT
AT
CG
GG
AA
TG
CT
GC
AA
AA
GC
GA
AA
AA
CG
G
GG
TA
AT
GG
CC
CT
GA
TA
TC
CC
TC
TA
AA
CA
TG
AT
TG
TC
GC
TG
TC
CA
GG
TG
AG
CA
GA
ÀG
A11
TF
T
TfC
GG
CK
fCT
GA
AT
CT
CA
C
CG
GC
AT
CG
TC
AA
CA
TA
AC
C
CA
TC
AA
GC
TG
TT
TG
TT
CG
TC
CG
CC
G
GT
TG
AA
CT
GA
GT
C1
7A
AT
AC
CA
GT
CT
CC
GG
AG
AA
CT
GG
GT
GC
AT
CA
C
GA
CG
GC
TG
TA
CT
GC
AG
GG
TG
TG
GC
G
TG
TT
GA
AA
GA
TC
AG
TC
CT
CA
AA
GT
GT
CG
ATF
IET
AT
TG
GT
GT
A
TG
AA
AG
CC
AT
CA
CA
TA
CT
GC
TC
GG
AT
GA
GT
AT
GG
GC
AA
C
GA
TC
AC
GG
GC
CA
CG
CT
GT
GT
C
GG
GT
AA
AA
AC
TG
GC
CT
AÀ
AÀ
TT
G
AG
CA
AT
AG
TT
AA
TG
U’T
GA
GC
TA
AA
G
TG
TA
GA
flT
GA
CC
GC
CA
CC
TT
GT
flT
CA
GT
AA
TG
GT
CG
GG
CC
TC
AA
AG
GT
flG
AT
GT
AC
C
AT
GA
TC
TA
AC
CC
TC
GG
TC
TC
GT
GT
GC
GG
AT
GA
AG
TC
AG
GC
TG
GG
CA
GC
AA
AC
TG
AT
AA
CT
CT
TG
AG
CA
TF
CT
CC
GC
TA
AC
TG
AT
AA
T
CG
GA
GG
AG
TA
AT
TA
CA
AA
CC
TG
GC
AT
AT
CC
iTT
CT
GC
AG
GG
AT
GC
AA
TA
CT
GC
GT
AG
AT
AT
TG
GC
TG
AG
CC
AT
CC
GA
TG
TC
AG
TT
TT
CT
G
DQ
D
Tri
met
hopr
im
Sul
fona
mid
es
Vir
ulen
cege
nes
99
543
X53
796
Pas
teur
Inst
itute
286
X07
848
I.A
.M
urra
y
215
Af2
5285
5D
.G
.W
hite
391
X00
926
J.H
arel
432
X12
868
O.
Skôl
d
265
X58
425
O.
SkiM
d
462
X57
730
C.
Wal
len
294
Z50
802
P.V
.A
dria
n
822
X12
869
R.
C.
Lev
esqu
e
722
M36
657
J.H
arel
793
X76
688
J.H
arel
351
AF0
7290
0J.
Har
el
408
X16
664
JHar
el
318
X61
239
J.H
arel
500
M10
133
J.H
arel
778
M18
96$
J.H
arel
Cla
ssI
inte
gron
qac
Et1
-su
!]A
TC
GC
AA
TA
GT
TG
GC
GA
AG
TG
CA
AG
GC
GG
AA
AC
CC
GC
GC
C79
7X
l287
0J.
Har
el
VR
GG
CA
TC
CA
AG
CA
GC
AA
GC
AA
GC
AG
AC
TT
GA
CC
TG
AT
Var
iabl
eX
l287
0
Ph
ylo
gen
itic
chuA
GA
CG
AÀ
CC
AA
CG
GT
CA
GG
AT
TG
CC
GC
CA
GT
AC
CA
AA
GA
CA
279
Af2
8039
6
gro
up
syj
aAT
GA
AG
TG
TC
AG
GA
GA
CG
CT
GA
TG
GA
GA
AT
GC
Gfl
CC
TC
AA
C21
1N
C_0
009
13
Tsp
E4
.C2
GA
GT
AA
TG
TC
GG
GG
CA
rTC
AC
GC
GC
CA
AC
AA
AG
TA
TIA
CG
152
AF2
2218
8
aA
lter
nati
veno
men
clat
ure.
DD 10
0
TA
BL
EII
.N
um
ber
ofre
sist
ant
anim
alan
dhum
anis
olat
esdet
erm
ined
bydi
scdi
ffus
ion
met
ho
d
No.
ofre
sist
ant
anim
alan
dhu
man
stra
ins
(%)
Ani
mal
isol
ates
AT
Ma
(n=
39)
Hum
anis
olat
esH
uman
isol
ates
Tot
alhu
man
from
1990
from
2001
isol
ates
(n=
16
)(n
54)
(n=
70)
AM
P2$
(72)
10(6
2)44
(81)
54(7
7)
CT
X,
XL
N7
(18)
0(0
)0
(0)
0(0
)
GE
N6
(15
)0
(0)
9(1
7)
9(1
3)
NE
O7
(18
)0
(0)
4(7
)4(6
)
KA
N8
(21
)0
(0)
7(1
3)
7(1
0)
lET
32(8
2)6
(38)
29(5
4)34
(48)
CH
L8
(21
)0
(0)
17(3
1)17
(24)
TM
P16
(41)
4(2
5)23
(42)
27(3
8)
SU
L30
(77)
$(5
0)40
(74)
48(6
$)
aAT
M,
anti
mic
robi
als:
AM
P,
ampi
cill
in;
CT
X,
cefo
taxi
me;
XL
N,
ceff
iofu
r;G
EN
,ge
ntam
icin
;N
EO
,
neom
ycin
;K
AN
,ka
nam
ycin
;C
HL
,ch
lora
mph
enic
ol;
TE
T,
tetr
acyc
line
;T
MP
,tr
imet
hopr
im;
SUL
,
sulf
onam
ides
Num
ber
ofpo
siti
veis
olat
esac
cord
ing
toth
eis
olat
ion
orig
in(%
)a
Ant
imic
robi
alag
ent
Gen
etic
mar
ker
prob
e
tet(
A)
tet(
B)
tet(
C)
tet(
D)
Ani
mal
isol
ates
17(5
3)
14(4
4)
15(4
7)
0(0
)
Hum
anis
olat
esfr
om
1990
2(3
3)
3(50
)
2(33
)
2(33
)
Hum
anis
olat
esfr
om
2001
13(4
6)
15(5
4)
10(3
6)
0(0
)
Tot
alhu
man
isol
ates
15(4
4)
18(5
3)
12(3
5)
2(6
)
DO
D 101
TA
BL
EII
I.A
nim
alan
dh
um
anis
olat
esw
hich
gave
apo
siti
veh
yb
rid
isat
ion
wit
hth
ere
sist
ance
and
viru
lenc
ete
sted
prob
ean
dha
dth
e
asso
ciat
edre
sist
ance
phen
oty
pe
Am
pici
llin
19(6
8)8
($0)
42(8
9)50
(8$)
blas
Hv
0(0
)3(
30)
4(9
)7
02
)
blO
OX
Ai
301)
0(0
)0
(0)
0(0
)
blao
xA7,
blap
sE4,
0(0
)0
(0)
0(0
)0
(0)
blac
TX
M3
Gen
tam
icin
aac(
3)-
lIa
1(1
7)0
(0)
3(3
3)3
(33)
aac(
3)-
H0
(0)
0(0
)0
(0)
0(0
)
ant(
2‘9
-la
Kan
amyc
inap
h(39
-Ia
6(8
6)0
(0)
2(5
0)2
(50)
ant(
2’9-
Ia0
(0)
0(0
)0
(0)
0(0
)
aph(
39
-lia
Neo
myc
inap
h(3
9-la
8(1
00)
0(0
)5
(71)
5(7
1)
ant(
2’9-
Ia0
(0)
0(0
)0
(0)
0(0
)
aph(
39
-lia
Tet
racy
clin
e
DD
D 102
tet(
E),
tet(
Y)
0(0
)0
(0)
0(0
)0
(0)
Chl
oram
phen
icol
catI
1(13
)0
(0)
14(8
2)14
(82)
catl
lO
(0)
0(0
)0
(0)
0(0
)
catI
il1(
13)
0(0
)0
(0)
0(0
)
foR
5(63
)0
(0)
1(6)
1(6)
Tri
met
hopr
imdh
frl
4(2
5)3
(75)
7(3
0)10
(37)
dI1f
rV6(
38)
0(0
)1(
4)1(
4)
dhfr
VII
0(0
)0
(0)
9(39
)9(
33)
dhfr
XII
I6
(38)
O(0
)1(
4)1(
4)
dhfr
lX0(
0)0
(0)
0(0
)0(0
)
Sul
fona
mid
essu
lI14
(50)
3(3
8)30
(75)
33(6
9)
sutI
l18
(67)
5(6
3)24
(60)
29(6
0)
Vir
ulen
ceaf
a31(
2)3
(19)
4(7
)7
(10)
afaD
85
(13)
0(0
)1(
2)1(
1)
sfaD
E10
(26)
8(50
)19
(35)
27(3
8)
pap
C19
(49)
12(7
5)25
(46)
37(5
3)
hlyA
9(2
3)11
(69)
23(4
3)34
(48)
iucD
30(7
7)4
(25)
30(5
6)34
(48)
Per
cent
age
ofpr
obe-
posi
tive
isol
ates
amon
gre
sist
ant
isol
ates
DD
D 103
TA
BL
EIV
.D
istr
ibuti
on
ofcl
ass
1in
tegro
nam
ong
the
anim
alan
dhum
anE
.co
ust
rain
s
No.
of
clas
s1
inte
gron
posi
tive
stra
ins
acco
rdin
gto
the
Isol
atio
nT
otal
no.
ofN
o.o
fcl
ass
1in
tegr
onle
ngth
ofth
eva
riab
lere
gion
(%)
peri
ods
stra
ins
posi
tive
stra
ins
(%)
0.5
kb1.
0kb
1.6
kb1.
7kb
1.8
kb
Ani
mal
3913
(33)
1(7
)8
(58)
3(2
1)1
(7)
0(0
)is
olat
es
Hum
an
isol
ates
163
(19)
0(0
)3
(100
)0
(0)
0(0
)0
(0)
from
1990
Hum
an
isol
ates
5419
(35)
1(5
)10
(53)
6(3
2)0
(0)
2(1
0)
from
2001
Tot
al
hum
an70
22(3
1)1
(4)
13(5
9)6
(2$)
0(0
)2
(9)
isol
ates T
otal
109
57(5
2)3
(5)
34(5
9)15
(26)
1(2
)4
(7)
Ant
imic
robi
alre
sist
ance
gene
s,cl
ass
1in
tegr
onan
dvi
rule
nce
gene
sa
tem
foR
aph(
32-
ladh
rfl
dh
rff
dhfr
XII
lsu
lIsu
llI
tet(
A)
tet(
B)
tet(
C)
Int
1af
aD8
sfaD
Epap
Cht
yA
foR
-
aph
(39
-!a
-+
dhfr
l-
--
dhfr
V+
+-
--
dhfr
XII
I-
--
--
su!!
--
--
-
su!!
!-
-+
+-
+-
-
tet(
A)
+-
--
--
--
tet(
B)
——
——
——
—(÷
)b
tet(
C)
+-
--
+-
--
++
+-
Inti
--
--
--
++
+-
+-
+
afaD
8-
--
--
+-
+-
+-
-
sfaD
E-
--
--
--
--
--
-
pap
C-
--
--
--
--
--
--
+
h!yA
--
--
--
--
--
--
-+
-4-+
iucD
--
(++)
(++)
ily
anti
mic
robi
alre
sist
ance
gene
sth
atex
hibi
ted
asso
ciat
ion
wit
han
othe
rge
nes
atth
eP
<0.
05le
vel
are
show
n.Si
gnff
ican
cele
vel
of
asso
ciat
ion
(as
asse
ssed
byth
ech
i-sq
uare
exac
tte
st):
-,P
>0.
05;
+,
0.05
?P
?0.
01;
-H-,
0.01
?F
?:0.
001;
+-H
-,0.
001
?F
.
bP
aren
thes
esin
dica
teth
ene
gati
veas
soci
atio
ns.
DD
D 104
TA
BL
EV
.A
ssoc
iati
onbe
twee
nan
tim
icro
bia
lre
sist
ance
gene
s,cl
ass
1in
tegro
nan
dv
iru
len
cege
nes
amon
gth
eE
xPE
Cst
rain
sis
olat
ed
from
anim
aIs
Ant
imic
robi
alre
sist
ance
gene
s,cl
ass
1in
tegr
onan
dvi
rule
nce
gene
sa
tem
shv
catI
aac(
3)-I
Idh
rfl
d1fr
Vil
sulI
suti
lte
t(A
)te
t(B
)te
t(C
)te
t(D
)In
t]sf
aDE
pap
Chl
yA
sh’.’
—
catI
--
aac(
3)-l
ia-
--
dI!f
rI-
--
-
d/f
rV
II-
--
++
-
sulI
--
--
-+
sulI
l-H
--1-
--
--
--
tet(
A)
--
--
--
-H-
-
tet(
B)
--
++
--
--
++
-
tet(
C)
--
--
--
++
-+
++
-
tet(
D)
-+
+-
-+
--
--
-
inti
--
--
-+
++
+-
+-
--
sfaD
E-
--
--
--
--
-+
--
pap
C-
--
--
--
--
--
--
+
htyA
--
--
--
--
-+
++
++
+
iucD
--
++
-+
++
-+
Onl
yan
tim
icro
bial
resi
stan
cege
nes
that
exhi
bite
das
soci
atio
nw
ithan
othe
rge
nes
atth
eF
<0
.05
leve
lar
esh
own.
Sig
nifi
canc
ele
vel
of
asso
ciat
ion
(as
asse
ssed
byth
ech
i-sq
uare
exac
ttes
t):
-,P
>0.
05;
+,
0.05
F?
0.01
;-H
-,0.
01?
P?
0.00
1;+
++
,0.
001
?F
.
++
DD 10
5
TA
BL
EV
I.A
ssoc
iati
onbe
twee
nan
tim
icro
bia
lre
sist
ance
gene
s,cl
ass
1in
tegro
nan
dvi
rule
nce
gene
sam
ong
the
ExP
EC
stra
ins
isol
ated
from
hum
ans
n,
SECTION III - Discussion
n
n
107
Discussion
Étude portant sur les isolats Escherichia cou entérotoxinogènes
Les bactéries Escherichia cou entérotoxinogènes (ETEC) sont des pathogènes du porc
causant des diarrhées chez les nouveau-nés et chez les porcelets en postsevrage. Les ETEC de
sérogroupe 0149: K91 sont plus souvent isolées depuis quelques années et représentent le
sérogroupe prédominant associé aux diarrhées chez le porc (Bischoff K. M. et al. 2002;
Teshager, T. et al. 2000; Mathew, A. G. et aÏ. 1999). Les ETEC sont aussi associées à certains
autres sérogroupes : 08, 09, 020, 0101, 0138 et 0141 (Garabal, J. I. et al. 1996). Malgré ce
nombre limité de sérogroupes, les souches ETEC montrent une certaine diversité génétique
(Nagy, B. and Fekete, p. z. 1999). En effet, par les évènements de transferts horizontaux, les
éléments génétiques acquis diffèrent entre les souches. Plusieurs clones semblent avoir évolué
en prenant différentes directions et en acquérant de nouveaux gènes.
Chez les animaux, les infections entériques deviennent difficiles à traiter puisque, de plus en
plus, l’agent pathogène en cause est résistant aux agents antimicrobiens. Ceci est dû, entre
autre, à la grande quantité d’agents antimicrobiens qu’ils reçoivent. En effet, certains groupes
d’animaux de ferme reçoivent des agents antimicrobiens dans leur moulé, en plus de ceux
utilisés pour le contrôle des infections. Même si les agents antimicrobiens sont employés
relativement fréquemment dans l’élevage des animaux et que la résistance à ceux-ci augmente
rapidement, peu d’études ont tenté de caractériser directement les gènes codant pour cette
résistance chez E. cou. Les études se sont plutôt intéressées au phénotype de résistance des
isolats. Très peu d’études donnent des informations sur les gènes impliqués dans la résistance
de ces isolats.
La caractérisation génétique de la résistance chez E. cou isolé d’animaux n’a été abordée, à
notre connaissance, que par deux récentes études. La première s’est penchée sur la
caractérisation génotypique de la résistance à la tétracycline, à la streptomycine et aux
sulfamides chez des souches E. cou isolées d’animaux et d’humains (Lanz, R. et al. 2003). La
deuxième s’est plutôt concentrée sur la caractérisation des profils de résistance aux
aminoglycosides retrouvés chez des souches E. cou d’origine porcine et bovine (Sandvang, D.
and Aarestrup, F. M. 2000). Notre étude démontre, quant à elle, les différents profils de
108
résistance en regroupant les 28 gènes de résistance différents retrouvés parmi 118 isolats
ETEC de sérogroupe 0149: K91 provenant de porcelets atteints d’une diarrhée postsevrage.
En plus d’offrir une information sur le génotype de résistance aux agents antimicrobiens chez
les ETEC, notre étude donne une information supplémentaire sur le changement des profils de
résistance qui se produit dans le temps puisque ces isolats ont été isolés sur une période de 23
années.
1.0 Profil phénotypique de la résistance
1.0.1 Sélection de la résistance
L’usage d’agents antimicrobiens permet une sélection des souches résistantes à ces agents.
Par exemple, l’utilisation d’ampicilline permet une sélection au niveau des souches résistantes
à l’ampicilline et ainsi, permet d’augmenter la concentration en bactéries résistantes à cet
agent. Dans notre étude, nous avons pu mettre en évidence cette notion en regardant les
agents antimicrobiens utilisés chez le porc et le phénotype de résistance exprimé par les
souches ETEC. Nous savons que les tétracyclines sont employées dans l’élevage du porc pour
le traitement d’infections de toutes sortes, mais aussi comme facteur de croissance et en
prophylaxie, depuis plusieurs années (Tableau I, Recension de la littérature). Or, chez les
isolats possédant au moins une résistance, nous observons que 90% sont résistantes à la
tétracycline. Depuis 197$ jusqu’en l’an 2000, le pourcentage de souches résistantes à cet
agent est demeuré assez stable. Les tétracyclines ont été utilisées de façon intensive avant
1978. Il n’est donc pas surprenant d’observer une résistance élevée. Toutefois, il ne faut pas
perdre en tête le fait que les isolats choisis pour cette étude ont à prime à bord été sélectionnés
à partir d’une banque de souches selon leur résistance. En effet, les souches ayant au moins
une résistance aux agents testés ont été sélectionnées pour cette étude. Quant au phénotype dc
résistance aux sulfamides, le même scénario est observé. Les isolats sont résistants à cet agent
dans 86% des cas et ce profil est resté stable depuis 1978.
Lorsque nous nous penchons sur le phénomène de résistance aux agents antimicrobiens
utilisés seulement en traitement, le profil est différent de celui observé pour ceux utilisés
comme facteur de croissance et/ou en prophylaxie. C’est le cas de l’ampicilline, de la
gentamicine et de la triméthoprime (Tableau I, Recension de la littérature) qui ne sont
administrés qu’en cas d’infections. L’ampicilline est donnée rarement chez le porc. Toutefois,
un homologue à l’ampicilline, l’amoxycilline, est employée fréquemment dans l’eau contre
109
les infections à Streptococcus suis et Haernophiltts parasuis. L’amoxycilline est une molécule
non homologuée en médecine vétérinaire qui se différencie de l’ampicilline par un
groupement OH au noyau benzène de la chaîne carboxyle. Cette substitution permet à
l’amoxycilline d’être plus soluble et plus facilement absorbée que sont homologue (Prescott,
J. f. et al. 2000). Cependant, la résistance bactérienne à l’un permet la résistance à l’autre
sous forme de résistance croisée.
La gentamicine semble encore être peu employée et lorsqu’elle est utilisée, elle est
administrée en injection afin de traiter différentes infections ches les porcs. La triméthoprime
peut être utilisée seule dans de rares cas mais, dans la majorité des cas, elle est utilisée en
combinaison avec les sulfamides dans l’eau ou en injection chez les porcs. On observe une
résistance à l’ampicilline, à la gentamicine et à la triméthoprime chez environ 25% des
souches E. cou. Même si la triméthoprime est normalement administrée en combinaison avec
une sulfamide, nous retrouvons un pourcentage d’isolats résistants à la triméthoprime moindre
que celui de la résistance aux sulfamides. L’utilisation des sulfamides seules dans la moulé
des porcs, comme facteur de croissance peut sélectionner les souches résistantes aux
sulfamides seulement et ainsi expliquerait cet écart. En raison des problèmes de résidus de
sulfamides dans les carcasses de viande, cet agent est de moins en moins employé chez les
porcs destinés à la consommation humaine. Par l’utilisation de facteurs de croissance, on crée
une situation où les souches résistantes sont sélectionnées, ce qui favorise le développement
de la multirésistance bactérienne. Ceci vient donc renforcer l’hypothèse que la résistance se
développe, non par une utilisation sporadique de forte quantité d’agents antimicrobiens, mais
plutôt par une utilisation soutenue.
La hausse de la résistance au chloramphénicol est un phénomène qui a attiré notre attention
puisque cet agent n’est plus employé au Canada depuis 1980 (Gilmore, A. 1986). La hausse
de la résistance pourrait s’expliquer par une sélection croisée produite par l’utilisation de
d’autres agents antimicrobiens ou par une utilisation illégale du chloramphénicol dans
l’élevage du porc. Étant donné que l’Agence Canadienne d’Inspection des Aliments (ACIA)
ne semble pas avoir retrouvé de traces de résidus de chloramphénicol dans les carcasses de
porc examinées depuis son bannissement, nous pouvons présumer que l’utilisation illégale de
chloramphénicol serait très sporadique si elle se faisait. La persistance ou l’augmentation de
la résistance au chloramphénicol chez des souches E. cou isolées d’animaux a aussi été
observée par d’autres auteurs (Bischoff, K. M. et al. 2002; Teshager, T. et aÏ. 2000; Lanz, R.
110
et al. 2003; Werckenthin, C. et al. 2002). Bourque et aï. ont aussi constaté durant leur étude
que la résistance au chloramphénicol ainsi qu’à la néomycine et à l’ampicilline, chez des
isolats ETEC, augmentait de façon significative lorsque les porcs recevaient dans leur moulé
100 ppm de néomycine et 100 ppm de tétracycline (Bourque et aÏ., 1980). Toutefois, il faut
noter que l’usage du chloramphénicol n’a pas été interdit dans tous les pays en même temps.
1.0.2 Émergence de la multirésistance
Lorsque nous observons de près nos résultats en rapport avec la phénotypie de la résistance
des isolats ETEC, nous pouvons voir une augmentation dans le temps de la multirésistance
(figure 2, Annexes). Les isolats de la première période (1978-1984) étaient résistants en
majorité (41%) à deux agents antimicrobiens testés et cette observation s’applique aussi pour
la deuxième période (1985-1989) (57%). Nous pouvons noter tout de même que certains
isolats dans ces deux périodes étaient multirésistants. Toutefois, à partir de 1990, le
pourcentage d’isolats multirésistants devient plus élevé pour atteindre 76% dans la dernière
période (1995-2000). Il semblerait que la résistance aux différents agents testés ait augmenté
rapidement à partir de 1995. Vers les années 90, nous avons pu être témoin d’une nouvelle
gestion des agents antimicrobiens dans l’élevage porcin, au Québec. Quarante pour cent des
porcelets au Québec sont sevrés entre 16 et 19 jours d’âge. Ce sevrage plus hâtif des porcelets
a contribué à l’établissement d’infections chez les porcelets en postsevrage et donc à
l’administration d’agents antimicrobiens en prophylaxie (communication personnelle). Il en
résulte une augmentation de la quantité d’agents antimicrobiens utilisés (chlortétracycline et
triméthoprime! sulfisoxasole) chez le porc afin de prévenir les infections à Streptococcus suis
ou à Haernophilus parasuis associées souvent au sevrage hâtif. Ceci expliquerait peut-être
l’augmentation de la multirésistance chez E. cou isolé de porcs. La multirésistance chez des
isolats E. cou provenant de porcs n’est pas rare. En effet, plusieurs études ont démontré la
résistance à plus de deux agents antimicrobiens chez plusieurs isolats E. cou (Amezcua, R. et
aï. 2002; Teshager, T. et aÏ. 2000; Mathew, A. G. et al. 1999; Lanz, R. et al. 2003; Sunde, M.
and Sorum, H. 1999; Mathew, A. G. et al. 199$). Les agents antimicrobiens les plus souvent
en cause sont la tétracycline, les sulfamides, la triméthoprime et le chloramphénicol. Nous
devons préciser que peut-être des résistances à d’autres aminoglycosides tels que la
streptomycine et la spectinomycine auraient pu être démontrer si nous avions testé ceux-ci.
Cependant, comme nous recherchions des gènes acquis, il nous apparaissait difficile
d’incorporer ces deux derniers agents puisque la résistance par mutation chromosomique peut
survenir facilement.
111
Notre étude surestime le nombre de souches résistantes, car seulement les souches démontrant
une résistance à un agent antimicrobien lors de leur isolement ont été analysées. Nous devons
noter que quelques souches, qui avaient présenté une ou des résistances à certains agents
antimicrobiens lors de leur caractérisation au Laboratoire E. cou de la faculté de médecine
vétérinaire, n’étaient plus résistantes aux agents testés par la suite. Ceci pourrait s’expliquer
par la perte de la résistance suite à des repiquages; un tel phénomène est reconnu depuis
longtemps en bactériologie.
1.1 Profil génotypique de la résistance
1.1.1 Diversité des gènes
Les résultats de la résistance phénotypique retrouvée chez les souches ETEC de notre
collection démontrent qu’il y a eu une augmentation de la multirésistance parmi les souches
résistantes dans le temps. La caractérisation de cette résistance a démontré qu’elle était due à
plusieurs gènes de résistance différents. Par exemple, la diversité génétique s’est beaucoup
accrue au niveau des gènes de résistance à la triméthoprime (Table 3, Article I). En effet, sur
les six gènes testés, quatre ont été retrouvés parmi les souches ETEC et associés une
résistance phénotypique à cet agent. Toutefois, nous savons qu’il existe plusieurs gènes de
résistance à la triméthoprime qui ont été décrits chez E. cou (Huovinen, P. et aÏ. 1995). Cette
diversité est donc déjà connue.
En plus de la diversité des gènes de résistance à la triméthoprime, nous remarquons que, dans
l’ensemble, certains gènes sont plus présents dans une période d’isolement donnée tandis que
d’autres le sont moins (Table 3, Article I). Les gènes de résistance aux aminoglycosides, à la
tétracycline ou aux sulfamides sont les gènes pour lesquels ce phénomène est le plus
fortement constaté. Le gène aph(3 )-IIa devient plus fréquent chez les souches isolées dans la
dernière période (1995-2000). Le gène tet(B) prédomine durant les trois premières périodes
pour, en fin de compte, devenir aussi fréquent que les gènes tet(A) et tet(C) dans la dernière
période. Finalement, un changement s’est fait dans la prévalence des gènes de résistance aux
sulfamides, sulI et suÏIL dont le deuxième devient aussi fréquent que le premier dans la
dernière période. Tous ces changements sont surtout constatés dans la dernière période
d’isolement des souches, ce qui laisse présumer qu’entre 1990 et 2000 certains facteurs
auraient influencé une sélection de certains gènes de résistance. Afin d’expliquer
112
l’augmentation de la prévalence du gène aph(3 9-lia dans la dernière période et la baisse de la
fréquence du gène aph(3 9-la, nous ne pouvons que supposer que la sélection provient de
l’émergence d’un clone possédant le gène aph(3 9-lia puisque, entre les deux enzymes codées
par ces gènes, il n’existe pas de différences au niveau de leur mode d’action ni au niveau de
leur spécificité. Les deux gènes codent pour une enzyme ayant une activité de
phosphorylation et tous les deux confèrent une résistance à la néomycine ainsi qu’à la
kanamycine (Shaw, K. J. et aÏ. 1993). Certaines études ont déjà mis en évidence la
propagation de gènes de résistance entre différents hôtes par l’isolement de souches ayant le
gène de résistance (Hunter, J. E. et aÏ. 1994).
La résistance croisée peut expliquer que certains gènes soient présents chez certains isolats à
un moment donné. Toutefois, d’autres facteurs peuvent aussi être la cause de l’augmentation
de la présence des gènes tet(A), tet(C) et sulli dans la dernière période. Une étude concernant
trois différentes porcheries ayant une gestion différente des agents antimicrobiens a retrouvé
en prédominance les gènes tet(A) et tet(C) chez les animaux non exposés aux agents
antimicrobiens et en prédominance le gène tet(B) chez les porcs en ayant reçu (Lee, C. et al.
1993). Le mode d’action aussi bien que la spécificité de l’enzyme antimicrobien peuvent
exercer une pression sélective positive et contribuer à l’émergence de nouveau gène dans le
temps. Par exemple, les classes des déterminants de résistance à la tétracycline A, B, et C sont
tous des enzymes ayant une fonction de pompe, mais possédant une spécificité différente. La
plupart des gènes tet codant pour des pompes confèrent une résistance aux tétracyclines, mais
non à la minocyline ou à la glycylcycline. Cependant, le gène tet(B) code pour une protéine
ayant une fonction de pompe qui confère non seulement une résistance à la tétracycline, mais
aussi à la doxycycline et à la minocycline (Testa, R. T. et al. 1993). Cette spécificité
différente entre les gènes tet peut contribuer à la perte ou l’émergence de certaines classes de
ces déterminants. Nous posons l’hypothèse que pour la souche, à certains moments, avoir un
de ces déterminants plutôt que les autres pourrait donner un certain avantage à celle-ci, d’où
l’émergence de ce déterminant.
D’autre part, le changement dans la présence des différents gènes sul ne peut être expliqué par
un changement dans l’utilisation des agents antimicrobiens. Toutefois, entre les deux gènes, il
existe une sensibilité différente pour le substrat PABA. En effet, le gène sulli aurait la
capacité de discerner le PABA de l’inhibiteur, les sulfamides (Skold, 0. 2001). Bref, la
présence du gène sulIL contrairement à celle du gène sul] offrirait à la souche un mécanisme
113
de résistance plus efficace, sans toutefois, sous-estimer la possibilité que ce gène soit apparu
suite à la sélection d’une souche ayant d’autres résistances.
Outre cela, nous remarquons que les gènes de résistance aux différents agents antimicrobiens
les plus fréquents correspondent parfois à ceux retrouvés par d’autres études. Par exemple, la
résistance aux f3-lactamines est conférée, dans la majorité des cas, par la présence des
déterminants tem et shv. Les gènes oxa, pse et ctx-rn n’ont pas été retrouvés, ou très peu
souvent, parmi les souches testées. Chez les entérobactéries résistantes aux 13-lactamines, les
études démontrent que tem est le gène le plus souvent détecté suivi du gène shv (Pai, H. et al.
1999) tandis que les gènes oxa, pse et ctx-m sont plus rarement retrouvés. Parmi les gènes de
résistance à la gentamicine testés, un seuL domine durant toute l’étude. Ce gène nommé
aac(3)-IV a aussi été détecté parmi des isolats E. cou de provenance animale par Sandvang et
al. (Sandvang, D. and Aarestrup, F. M. 2000). Rappelons que ce gène confère, en plus d’une
résistance à la gentamicine, une résistance à l’apramycine, un agent antimicrobien autorisé en
médecine vétérinaire depuis 1988. Ces deux agents sont utilisés intensivement dans le
traitement de la colibacillose en postsevrage chez le porc (Amezcua, R. et al. 2002). Nous
remarquons que les deux autres gènes de résistance aux aminoglycosides observés parmi nos
souches sont tous les deux des gènes qui codent pour une phosphotransférase soit de type I ou
de type II (Table 3, Article I). Ces gènes donnent une résistance croisée à la néomycine et à la
kanamycine (34). Ceci explique le fait que nous ayons retrouvé des souches résistantes à la
kanamycine, et ce, même si cet agent antimicrobien n’est pas utilisé dans l’industrie porcine
au Canada. Contrairement à nos résultats, Sandvang et al. ont retrouvé les gènes de résistance
ant(2 ‘)-Ia et aac(3)-IIa comme étant les plus prévalant parmi des isolats de provenance
porcine (Sandvang, D. and Aarestrup, F. M. 2000). Toutefois, les gènes de résistance aux
aminoglycosides sont très nombreux à être caractérisés (voir Recension de littérature). Il n’est
donc pas tellement surprenant que, d’une étude à une autre, la prévalence de ces gènes diffère.
Une étude sur ces gènes de résistance, mais basée sur des isolats de provenance humaine, a
démontré que la fréquence de détection de ces gènes varie dans le temps et selon la région
géographique de l’isolement de la souche (Miller, G. H. et al. 1997). La résistance non
attendue au chloramphénicol semble provenir de la présence du gène catI. Les gènes catil et
catIlI n’ont pas été détectés parmi les souches ETEC. Une étude accomplie par Bischoff et al
(Bischoff K. M. et al. 2002), a démontré que parmi les 48 souches E. cou résistantes au
chlorampliénicol testées seulement quatre portaient le gène catI ; le gène relativement
méconnu, cmlA, était responsable de la résistance des autres souches. Ce gène cmlA a été
114
caractérisé par Bissonnette et al. (Bissonnette, L. et al. 1991) et code pour une résistance non
enzymatique due à la synthèse de pompe. Avant l’étude de Bischoff et al. en 2002, le gène
crnÏA avait été retrouvé que très rarement parmi des isolats E. cou. Peut-être ce gène, s’il avait
été testé dans notre étude, serait-il présent chez les quatre souches résistantes au
chloramphénicol, mais négatives à l’hybridation pour les gènes cat. Finalement, le gène foR
n’a été détecté que chez une seule souche. D’autres études ont démontré une émergence de ce
gène parmi les souches isolées d’animaux (White, D. G. et al. 2000; Cloeckaert, A. et al.
2000; Keyes, K. et al. 2000). Par contre, aucune de nos souches n’était résistante au
florfénicol qui n’est utilisé que depuis quelques années dans l’industrie du porc.
De façon analogue aux gènes de résistance aux aminoglycosides, plusieurs gènes de
résistance à la triméthoprime ont été décrits. Sur les six différents gènes testés, quatre ont été
détectés parmi les souches ETEC. Tous les quatre ont été isolés plus fréquemment dans la
dernière période (1995-2000) d’isolement des souches (Table 3, Article I). Nos résultats
peuvent être comparés à ceux obtenus dans une étude réalisée en Corée sur des isolats E. cou
de provenance d’infections urinaires chez les humains (Lee, J. C. et aÏ. 2001). Les auteurs de
cette étude ont pu démontrer la présence, par ordre d’importance décroissante, des gènes
dhfrXVIL dhfrX[L dhfrL dhfrV et dhfr VII chez leurs isolats. La présence des trois derniers a
été vérifiée dans notre étude, mais dhfr VII n’a pas été retrouvés. La résistance aux sulfamides,
qui est souvent associée à celle de la triméthoprime, est due, parmi les souches étudiées, à la
présence du gène sulI (en majorité) suivi du gène sulli. La prévalence de sulI parmi des
isolats de porc a aussi été remarquée par Lanz et al (Lanz, R. et al. 2003).
1.1.2 Mobilité des gènes de résistance
La haute fréquence du gène sulI parmi les souches ETEC nous amène à croire en la présence
d’intégrons de classe 1 chez ces souches. En effet, l’intégron de classe 1 est caractérisé par
une région 3’-conservée contenant les gènes qacEAl et sulI (figure 1, Recension de la
littérature). La présence d’intégrons permet à la souche d’insérer plusieurs gènes de résistance
aux agents antimicrobiens sous forme de cassettes à l’intérieur de celui-ci et donc, permet la
dissémination, en plus de l’accumulation, de ces gènes. Comparativement aux plasmides, qui
sont des éléments génétiques mobiles possédant une origine de réplication propre, l’intégron
ne peut pas se répliquer de façon autonome. Il doit alors être répliqué soit grâce à l’origine de
réplication d’un plasmide ou grâce à la réplication du chromosome de la bactérie hôte, selon
si l’intégron est situé sur l’un ou l’autre.
115
Parmi les souches ETEC testées, la majorité des souches (62%) possèdent un intégron de
classe 1 (Table 4, Article I). Une haute fréquence d’intégron de classe 1 parmi des souches E.
cou isolées d’animaux a aussi été notée par d’autres auteurs (Sandvang, D. and Aarestrup, F.
M. 2000; Lanz, R. et aÏ. 2003; Sunde, M. and Sorum, H. 1999; Bass, L. et al. 1999).
L’intégron de classe 1 représente donc un moyen efficace de dissémination de la résistance
employé par plusieurs souches E. cou. Parce que l’intégron est caractérisé par sa fonction
d’intégration de gènes à son site d’insertion situé dans la région variable de l’intégron (RV),
la caractérisation de cette région peut nous donner plusieurs informations sur les gènes de
résistance qu’il transporte ainsi que sur la résistance multiple que certaines souches
démontrent. À la suite de l’amplification de cette RV, nous avons obtenu quatre longueurs
d’amplicons différentes. Au total, ces quatre RV différentes comportent cinq différents gènes
de résistance intégrés. Parmi nos souches, les RV qui ont été le plus souvent amplifiées
mesurent 1,3 et 1,0-kb (Table 4, Article I). Les longueurs de RV amplifiées des intégrons
contenus chez des isolats E. coÏi producteurs d’une Shiga-toxine (STEC) étaient de longueurs
similaires aux nôtres (Zhao, S. et al. 2001). Il y a donc possiblement une dissémination de ces
intégrons parmi des souches E. cou de différents sérogroupes. De plus, ces mêmes longueurs
de RV d’intégrons de classe 1 ont aussi été amplifiées chez d’autres espèces bactériennes
(Kiebsiella pneumoniae, K. oxytoca, Enterobacter cloacae et E. aerogenes) (Schrnitz, F. J. et
aÏ. 2001; van Belkum, A. et al. 2001).
Puisque la longueur de la RV nous donne un indice sur les gènes contenus, nous pouvons
alors croire que les différentes RV démontreraient différents gènes intégrés. Nous pourrions
aussi croire que, plus la RV est longue, plus elle contient de gènes. Lors du séquençage de
deux représentants de longueurs différentes de RV, cinq différents gènes de résistance ont pu
être détectés. Il s’agit de gènes de résistance à la triméthoprime et aux sulfamides et des
résistances les plus souvent décrites résidant sur ce type (Bass, L. et al. 1999; Roy, P. H.
1997; Peters, E. D. et al. 2001). Ceci nous laisse supposer que les gènes de résistance aux 13-lactamines, à la tétracycline et au chloramphénicol ne sont pas associés avec les intégrons de
classe 1 présents chez les souches ETEC testées. De plus, les gènes contenus dans les
différentes RV correspondent avec ceux retrouvés sur des RV de même longueur que les
nôtres, mais isolées des souches STEC (Zhao, S. et al. 2001).
Il est intéressant de voir que les RV les plus longues sont plus fréquentes dans la dernière
période d’isolement des souches. Ce phénomène a aussi été remarqué chez des souches
116
isolées de spécimens de sang entre les années 1993 et 1999 ($chmitz, F. J. et al. 2001). Ce qui
laisse croire que, dans le temps, il y aurait une accumulation de cassettes dans la RV, ce qui
augmenterait progressivement la longueur de la RV. Toutefois, le nombre de cassette ne
dépasse pas deux dans notre étude, ce qui limite l’application de cette hypothèse dans notre
cas. Notre étude montre que les souches possédant les mêmes longueurs de RV amplifiées
semblent avoir les mêmes gènes de résistance intégrés. Cependant, aucun des gènes de
résistance ciblés par l’hybridation sur colonies n’a été retrouvé parmi les RV séquencées.
1.2 Association des gènes de résistance et co-sélection
Lors de l’analyse des résultats de nos hybridations sur colonies, nous avons pu mettre en
évidence certaines associations de gènes de résistance. Il a été statistiquement prouvé que
certains gènes sont associés de façon positive ou négative, selon la présence ensemble ou non
de ceux-ci dans l’isolat. Ainsi, les gènes tem et shv ont montré une association très forte et
positive entre eux (Table 5, Article I). Le gène tem a été décrit pour la première fois par
Richmond en 1975, tandis que le gène shv a été décrit pour la première fois par Nugent en
1979. En prenant pour acquis que les bactéries accumulent les gènes de résistance si cela leur
est favorable, nous pourrions expliquer cette association en posant l’hypothèse que toutes les
souches qui possèdent le gène shv posséderaient déjà au départ le gène tem puisque celui-ci
est présent depuis plus longtemps que shv dans l’environnement. Une seconde association a
attiré notre attention, il s’agit des gènes tet(A)/tet(C) qui sont tous les deux très fortement
associés positivement entre eux et négativement avec tet(B). Ces résultats ne peuvent être dus
à l’homologie de séquence de ces deux gènes car une technique plus sensible que
l’hybridation, la PCR, nous à permis de vérifier la présence de ces deux gènes chez les
souches. Nous savons que les déterminants tet sont très souvent associés à des plasmides. Une
étude menée par Jones et al. (Jones, C. S. et al. 1992) a démontré que la plupart de ces
plasmides avaient des incompatibilités entre eux et donc ne pouvaient se retrouver dans la
même souche ensemble. Nous pourrions donc expliquer ces associations en supposant que les
gènes tet(A)/tet(C) sont situés sur des plasmides compatibles mais incompatibles avec celui
qui porte le gène tet(B) La présence des gènes tet(A) et tet(C) chez une même souche a aussi
été observée par Lee et aï. (Lee, C. et al. 1993).
En plus d’observer une association entre les différents gènes de résistance, nous pouvons
remarquer que ces associations changent dans le temps. En effet, certaines associations entre
117
les gènes peuvent être remarquées seulement lorsque les souches sont divisées en deux
groupes ; le premier groupe est composé des souches isolées entre 197$ et 1989 et le
deuxième groupe est composé des souches isolées entre 1990 et 2000 (Tableau III et IV,
Annexes). De la sorte, nous pouvons remarquer que l’association entre le gène tem et les
gènes tet, entre le gène aph(3 )-Ia et tet(A)/tet(C) ainsi que l’association entre suli et catI ne
sont présentes que chez les souches isolées dans le premier groupe. Toutefois d’autres
associations présentes uniquement dans le deuxième groupe ont été notées. Par exemple,
l’association entre les gènes acc(3)-IV, dhfr V et catI a été observée. Il est donc intéressant de
souligner ici que le gène de résistance au chloramphénicol, catI, est associé dans le premier
groupe avec le gène de résistance aux sulfamides tandis que, dans le deuxième groupe, il est
plutôt associé avec les gènes de résistance à la gentamicine et à la triméthoprime. Puisque le
chloramphénicol n’est plus utilisé en production porcine au Canada depuis 1980 et qu’il
semblerait que la résistance à cet agent continue d’exister, nous pouvons croire en une co
sélection de cette résistance. L’utilisation de sulfamide, de gentamicine ou de triméthoprime
pourrait sélectionner des souches résistantes au chloramphénicol grâce à l’association qui a
été démontrée entre les gènes de résistance à ces différents agents antimicrobiens et au gène
cati.
1.3 Changements factoriels et diversité des profils
Différents profils phénotypiques et génétiques de la résistance ont été déterminés parmi les
souches étudiées. Certains facteurs peuvent être à l’origine de ces changements. Un de ces
facteurs, qui a d’ailleurs été mis en évidence tout au long de l’étude, est le temps. L’année
d’isolement des souches peut influencer le profil de résistance de la souche. En effet, certaines
résistances phénotypiques se sont développées dans le temps. La résistance à la gentamicine,
chloramphénicol et à la tnméthoprime en sont des exemples (Table 2, Article I). Toutefois,
même si nous remarquons qu’il y a peu de variation au niveau du pourcentage de souche
résistante à un agent antimicrobien donné dans le temps, les gènes qui sont impliqués dans la
résistance à celui-ci peuvent être différents selon l’année d’isolement de la souche. L’exemple
des gènes tet et sut (voir deuxième partie de la discussion; mobilité des gènes) démontre bien
ce phénomène.
Cependant, le facteur temps n’est pas le seul à intervenir dans le changement des profils
observés. Le mode d’action de l’enzyme codée par le gène de résistance ainsi que sa
11$
spécificité pour le substrat sont des éléments qui ont été démontrés dans la deuxième partie de
cette discussion. Le site géographique d’isolement de la souche, le dosage et la gestion des
agents antimicrobiens ainsi que le stade de la croissance du porc infecté peuvent influencer les
résistances développées par les souches et donc leur profil de résistance. Les souches
proviennent de porcs élevés dans la province de Québec. La région géographique est donc
grande et les vétérinaires qui y utilisent des agents antimicrobiens sont nombreux. Nous
savons que le fait d’administrer de façon parentérale des agents antimicrobiens, en plus des
autres agents donnés dans la moulé, augmente la résistance à d’autres agents, due à une co
sélection des résistances. Par exemple, il a été montré que l’injection d’oxytétracycline chez
les bovins qui reçoivent dans leur moulé de la chlortétracycline est associée à l’augmentation
de la prévalence de la résistance au chloramphénicol et au sulfisoxasole (OConnor, A. M. et
al. 2002). Moins de souches multirésistantes sont observées chez les fermes qui utilisent peu
d’agents antimicrobiens (Mathew, A. G. et al. 1999; Mathew, A. G. et aÏ. 1998).
Un dernier facteur mérite d’être considéré. Il s’agit de l’âge des porcs desquels proviennent
les souches ETEC. Plus de 97% des porcs de notre étude étaient âgés entre 1 et 70 jours
d’âge. Nous remarquons que, lorsque la souche ETEC est isolée d’un porc âgé entre 1 et 10
jours, celle-ci est en général (40%) résistante à deux agents antimicrobiens (figure 3,
Annexes). En bas de 10 jours, les porcelets ne reçoivent pas d’agents antimicrobiens. Par la
suite, entre 10 et 60 jours, les porcs reçoivent des agents antimicrobiens dans leur moulé et en
traitement. Il n’est donc pas surprenant d’observer que plus l’âge du porc est élevé, plus les
isolats sont résistants à plusieurs agents. Ce phénomène est surtout remarqué en bas de 30
jours d’âge. Par la suite, entre 31 et 70 jours, le nombre de résistance des souches isolées est
très variable. Par contre, Mathew et al. ont aussi observé une augmentation de la
multirésistance chez les souches isolées de porcs âgés de plus de 30 jours (Mathew, A. G. et
al. 1998). Toutefois, il faut prendre en considération que, dans cette étude, la majorité des
souches isolées de porcs âgés entre 1 et 20 jours l’ont été entre 1978 et 1989 (figure 4,
Annexes). Entre 1995 et 2000, les souches testées proviennent en majorité de porcs âgés entre
21 et 70 jours. Nous savons que, dans le temps, une augmentation de la multirésistance a été
observée. Par contre, nous ne savons pas comment cela se présenterait si les souches isolées
dans la quatrième période (1995-2000) avaient été isolées en majorité de porcs âgés entre 1 et
20 jours. Le faible nombre de souches provenant de porcs âgés entre 1 et 20 jours entre les
années 1995 à 2000, est probablement dû au fait que, lorsqu’un porcelet âgé de moins de 20
jours est atteint d’une diarrhée, les chances que cela soit dû à une infection à E. colt sont très
119
élevées. Alors, les vétérinaires envoient que très rarement des échantiltons de ces porcs au
laboratoire pour un diagnostic et traitent le porcelet immédiatement avec des agents
antimicrobiens. Puisque notre collection est formée des échantillons reçus au laboratoire de
diagnostique de la faculté vétérinaire à Saint-Hyacinthe entre 1978 et 2000, il en résulte donc
un faible lot de souches isolées de porcelets âgés entre 1 et 20 jours entre les années 1995 et
2000.
1.4 Expériences et analyses à venir
Comme qu’il a été souligné dans la sixième partie de la recension de la littérature, les gènes
tem et shv possèdent de nombreux variants dus à des mutations ponctuelles dans le gène.
Celles-ci n’ont pas pu être identifiées par la méthode choisie dans cette étude. Toutefois, il
serait intéressant de les identifier afin de pouvoir les différencier d’une souche à l’autre. Ceci
nous permettrait d’affiner notre caractérisation des souches. Effectivement, nous savons
présentement que 35% des souches testées possédaient le gène tem, mais nous ne pouvons pas
déterminer s’il s’agit du même variant. Ceci pourrait être confirmé en faisant, autre que du
séquençage, du «Ligase Chain Reaction» (LCR) sur l’ADN des souches qui a comme
principe de permettre une élongation du brin cible seulement si la mutation ponctuelle choisie
est présente (Kim, J. and Lee, H. J. 2000; Barany, F. 1991). Puisque les variants du gène tem
et shv connus sont séquencés, il serait donc possible de faire ce genre d’expérience.
De plus, plusieurs souches possèdent certaines ressemblances au niveau des profils
génotypiques de résistance et certains gènes semblent être associés entre eux. Il semblerait
donc que certaines souches proviendraient d’un même clone ou du même ancêtre. Toutefois,
notre étude ne peut confirmer cette hypothèse. Une étude plus approfondie en ce qui a trait
aux profils de digestion de l’ADN de ces souches ETEC pourrait nous informer davantage sur
l’évolution de celles-ci. En effet, la migration en champs pulsés de l’ADN digéré par des
enzymes de restriction nous permet d’obtenir des empreintes propres à la souche. Ainsi, une
comparaison des ces profils permettrait de retracer certains clones et d’évaluer la distance
phylogénétique qui existe entre les souches analysées (Rasheed, J. K. et al. 1997). Par contre,
d’autres méthodes de typage peuvent aussi être utilisées comme l’analyse des plasmides ainsi
que l’analyse des résultats de PCR avec des amorces arbitraires (RAPD).
120
Étude sur les souches Escherichia cou pathogènes extraintestinales d’origine
animale et humaine
Il est connu que certaines souches E. cou peuvent causer des entéropathies. Toutefois,
certaines autres souches causent des infections extraintestinales chez leur hôte sans générer de
problème au niveau intestinal. Les ExPEC possèdent plusieurs similarités entre eux (facteurs
de virulence, profil phylogénétique,...) laissant croire à des populations de souches croisées
entre les différents hôtes (Jobnson, J. R. et al. 2000; Jolmson, J. R. et al. 2002). L’utilisation
des agents antimicrobiens est jusqu’à maintenant l’arme d’excellence afin de contrôler les
infections extraintestinales causées par les ExPEC, cependant, plusieurs souches sont
devenues résistantes à ceux-ci. La caractérisation des gènes de résistance chez les isolats
ExPEC permettrait d’établir des profils et de les comparer entre eux afin d’établir certains
liens entre les différentes souches. Or, très peu d’études se sont concentrées à faire une
caractérisation génotypique de la résistance chez les souches E. cou isolées de différents
hôtes.
Notre étude a permis d’établir différents profils de résistance regroupant 28 gènes de
résistance différents retrouvés parmi 39 isolats ExPEC isolées de tissus infectés chez des
animaux et 70 isolats ExPEC isolées d’urine provenant d’humains. Il est à noter que ces
isolats ont été sélectionnés sur la base de leur résistance à au moins un des agents
antimicrobiens. Puisqu’en médecine vétérinaire, peu d’échantillons d’urine provenant
d’animaux révèlent une résistance à un ou plusieurs agents antimicrobiens, nous avons donc
préféré élargir le choix des échantillons en acceptant des isolats ExPEC animaux provenant de
tissus infectés. Ce problème n’a pas existé lors de la collecte des échantillons d’urine de
provenance humaine mais aurait existé si nous avions préféré la collecte de souches provenant
de tissus infectés humains. Ces isolats ont été caractérisés selon leur phénotype de résistance à
10 agents antimicrobiens et selon la présence ou absence des différents gènes de résistance
choisis.
2.0 Profil phénotypique de la résistance
2.0.1 Vision d’ensemble
121
Les profils phénotypiques de résistance établis chez les isolats d’animaux et d’humains étaient
relativement semblables. De façon générale, les isolats étaient résistantes à l’ampicilline, à la
tétracycline et aux sulfamides (Table II, Article II). Ces agents sont utilisés en thérapie et/ou
comme facteurs de croissance et/ou donnés en prophylaxie aux animaux mais ils sont aussi
couramment prescrits aux humains (Tableau I, recension de littérature). Il n’est donc pas
surprenant de noter une fréquence élevée de la résistance à ces agents. Nous remarquons que
la résistance aux céphalosporines (cefotaxirne et cefliofur) n’est exprimée que par des isolats
provenant d’animaux. Ceci peut s’expliquer par le fait que ces agents sont fréquemment
administrés aux animaux destinés à la consommation humaine. Par contre, il semblerait que la
résistance aux céphalosporines chez l’animal soit plutôt rare et lorsqu’elle est présente, elle se
retrouve surtout parmi les isolats de provenance bovine et aviaire (Lanz, R. et al. 2003;
Chanal-Claris, C. et aÏ. 1997) se qui correspond aussi avec nos résultats. La kanamycine est
un aminoglycoside administré peu fréquemment à l’animal. Par contre, nous avons remarqué
un pourcentage plus élevé d’isolats de provenance animale résistants à cet agent
comparativement à ceux de provenance humaine. Dans ce cas, nous pouvons croire que cette
résistance est due à un phénomène de résistance croisée avec d’autres aminoglycosides,
particulièrement la néomycine (Recension de la littérature, sixième partie).
La comparaison des profils entre les isolats provenant des différents animaux a permis de
noter une variation importante entre ceux-ci. Les isolats provenant des animaux de compagnie
sont ceux qui possédaient le moins de résistance. Toutefois, ces isolats proviennent
d’échantillons d’urines, comparativement aux autres isolats qui proviennent en majorité de
tissus ou d’organes infectés. D’autres auteurs ont aussi remarqué que la multirésistance
bactérienne provenant d’isolats d’urine animale était peu fréquente (Oluoch, A. O. et al.
2001). Plusieurs différences au niveau du phénotype de résistance ont été observées entre les
isolais provenant d’urine humaine isolés en 1990 et ceux isolés en 2000. Effectivement, les
souches isolées en 1990 possèdent généralement deux fois moins de résistances que celles
isolées en 2000 (Table II, Article II). Une augmentation de la multirésistance dans le temps
est donc notée comme nous l’avons aussi relevé parmi les isolats ETEC dans la première
partie de la discussion de ce mémoire.
2.0.2 Homogénéité dans les profils de multirésistance
Dans notre étude, il n’a pas été rare de rencontrer des isolats qui possédaient plus de deux
résistances. Environ la moitié des isolats provenant d’animaux et d’humains avaient plus de
122
deux résistances. Aucun profil de multirésistance n’a pu être associé à l’origine de la souche
car nous retrouvons les mêmes résistances dans chacune des origines (animale et humaine).
Toutefois, certaines souches isolées d’animaux possédaient jusqu’à 10 résistances, ce qui n’a
pas pu être observé parmi les isolats de provenance humaine. Il semblerait donc que la
caractérisation de la résistance au niveau du phénotype ne serait pas suffisante afin d’établir
l’origine de l’isolat. Véritablement, les isolats ExPEC, peu importe leur origine d’isolement,
possèdent un profil phénotypique de résistance semblable. Ceci a aussi été observé par Lanz
et aï. parmi des souches E. cou isolées d’infections extraintestinales de différents animaux
(Lanz, R. et al. 2003). Par contre, lorsque ceux-ci ont comparé le phénotype de ces isolats
ExPEC à celui des souches E. cou isolées des fèces de porcs, ils ont noté une différence entre
les profils de résistance. Puisque les profils établis à partir du phénotype de résistance des
isolats ExPEC sont similaires, nous pourrions croire en la possibilité que ces groupes ne
forment pas deux groupes de souches distincts.
2.1 Profil génotypique de la résistance
2.1.1 Diversité des gènes
La comparaison des profils établis à partir du phénotype de résistance des différentes souches
ExPEC, isolées d’animaux et d’humains, n’a pas permis l’établissement de patrons
particuliers. Par contre, par l’hybridation sur colonies des différentes sondes représentant les
28 gènes de résistance choisis, nous observons que différents gènes sont acquis par les isolats
de provenance animale de ceux acquis par les isolats provenant d’humains. En effet, différents
profils établis à partir du génotype de résistance ont pu être précisés entre les différentes
origines (animale ou humaine)(Table III, Article II). Pour un même phénotype, des gènes
différents sont sollicités chez l’animal versus l’humain. Par exemple, le gène oxa-1 n’est
présent que chez les isolats de provenance animale tandis que le gène shv n’est présent quant
à lui, que chez les isolats de provenance humaine. À l’occasion, les mêmes gènes ont été
détectés pamiis les souches isolées chez les deux origines, mais à des fréquences variées. Par
exemple, les gènes catI sont plus fréquemment détectés chez les isolats d’origine humaine et
le gènefloR, chez les isolats d’origine animale.
En plus d’observer qu’il y a certains gènes qui sont plus souvent retrouvés parmi les souches
isolées des animaux ou des humains, certaines particularités ont aussi été observées entre les
isolats isolées de l’humain à des périodes différentes. Les souches isolées d’urine de
123
provenance humaine en 1990 résistantes à la tétracycline portent, dans 33% des cas, le gène
tet(D). Ce gène n’a pas été retrouvé parmi les isolats de provenance humaine isolés en 2000.
De plus, les souches en 1990 qui étaient résistantes aux sulfamides possédaient soit le gène
sutI ou le gène sutil, mais jamais les deux ensemble. Toutefois, il n’est pas rare de retrouver
ces deux gènes parmi les souches isolées en 2000. Ce qui laisse croire que les souches ont
évolué dans le temps et qu’en plus de posséder plusieurs résistances, les souches isolées en
2000, ont pu se transférer les gènes de résistance. Ceci démontre encore une fois la mobilité
ou l’instabilité des gènes de résistance.
Généralement, un gène de résistance a pu être associé à un phénotype de résistance chez
l’isolat. Toutefois, seulement deux des sept souches résistantes aux céphalosporines semblent
posséder un des gènes de résistance aux céphalosporines testés. Une possède le gène tem et
l’autre le gène oxa-1. Par contre, il n’est pas rare de retrouver parmi les isolats E. cou
résistantes aux «Extended-spectmm 13-Lactamases» (EB$L) une hyper-production d’une f3-
lactamase chromosomique ou plasmidique de type AmpC (Feria, C. et al. 2002; Bradford, P.
A. et aÏ. 1999) ce qui pourrait être la cause de la résistance aux céphalosporines des cinq
autres souches.
2.1.2 Mobilité des gènes
Plusieurs gènes de résistance semblent être distribués parmi les différents hôtes. Le transfert
de souches résistantes entre les différents hôtes ou bien, le transfert de matériel génétique
(gènes de résistance, intégrons, plasmides,...) entre souches d’hôtes différents ou non, sont là
deux phénomènes qui peuvent favoriser la distribution des gènes de résistance entre les
différents hôtes (Oppegaard, H. et al. 2001; Sunde, M. and Sorum, H. 1999; van den Bogaard,
A. E. et aÏ. 2001). La détection de l’intégron de classe 1 parmi les souches ExPEC et la
caractérisation de la RV de ceux-ci peut nous permettre d’orienter notre choix sur une des
deux causes suggérées ci-haut.
Comparativement aux échantillons ETEC étudiés dans cette étude, la présence d’intégron de
classe 1 parmi les isolats ExPEC de provenance animale et humaine est deux fois moins
fréquente (Table TV, Article II). Toutefois, ceux-ci sont impliqués dans la résistance aux
agents antimicrobiens de ces souches puisqu’après caractérisation de ces intégrons, cinq
différentes longueurs de RV ont été obtenues contenant différents gènes de résistance. Nous
retrouvons parmi les isolats animaux et humains ces différentes longueurs de RV. Elles
124
correspondent aux longueurs de RV généralement amplifiées (Sandvang, D. and Aarestrup, F.
M. 2000; Sunde, M. and Sorum, H. 1999; Lee, J. C. et al. 2001; Bass, L. et al. 1999). Trois de
ces longueurs ont aussi été retrouvées parmi les isolats ETEC et correspondent à celles les
plus fréquemment amplifiées soit 1,0, 1,3 et 1,6-kb.
Certaines différences entre les isolats de provenance animale ont été observées. En effet, les
isolats de porcs semblent posséder plus fréquemment des intégrons de classe 1. Cependant,
aucune longueur de RV n’est attribuable à un type d’isolat en particulier. Dans la littérature,
aucune étude n’a pu établir ce lien, même s’il existe certaines particularités entre les différents
genres bactériens (Schmitz, F. J. et al. 2001).
2.2 Association et co-sélection des gènes de résistance
Avoir des informations sur l’association entre les gènes de résistance, dans une population
donnée, permet d’évaluer les possibilités de co-sélection. Le gène sulI qui est associé à la
région conservée 3’ de l’intégron de classe 1, permet de croire que l’utilisation de sulfamides
permettrait à l’intégron de classe 1 et aux gènes intégrés dans la RV de celui-ci, d’être
conservés par la bactérie. Dans ce cas, l’usage d’agents antimicrobiens permet une co
sélection de plusieurs gènes de résistance et oriente la population bactérienne vers la
multirésistance (Bass, L. et al. 1999). Dans cette étude, nous avons démontré qu’il existait, en
plus de l’association entre sulI et l’intégron de classe 1, la présence d’associations entre les
gènes de résistance. Ces associations diffèrent selon l’origine de l’isolat. En effet, chez les
isolats de provenance animale, il existe une forte association (0.01 ? P 0.001) entre les
gènes tem et dhfr V (Table V, Article II). Toutefois, le gène tem parmi les isolats humains est
plutôt associé avec la présence du gène sulIl (0.00 1 ? P) (Table VI, Article II).
De plus, il semblerait que le profil d’association des gènes soit conservé selon la région
géographique de l’isolement des souches. Dans la province de Québec, l’association des
gènes de résistance à la tétracycline tet(A) et tet(C) semble être présente chez les isolats
animaux et humains (Article I et Article II). Toutefois, même si cette association a déjà été
rencontrée par d’autres auteurs, celle-ci n’a jamais été aussi forte dans les autres régions
étudiées (Lanz, R. et al. 2003; Lee, C. et al. 1993).
2.3 Virulence et phylogénie des souches ExPEC
125
2.3.1 Profil de virulence
Parmi les isolats ExPEC, l’opéron pap (pyelonephritis-associated pili), sfa (S fimbrial
adhesin) et afa (afimbrial adhesin) codent pour les adhésines les plus fréquemment isolées.
Ces adhésines permettent l’attachement spécifique de la bactérie aux différents récepteurs
situés à la surface de la cellule épithéliale (Smyth, C. J. et aÏ. 1994). Le gène pap a été détecté
parmi 49% des isolats animaux étudiés dans cette étude et 61% des isolats humains. Même si
la moitié des isolats possèdent seulement l’adhésine P codée par le gène pap, 22% des isolats
possèdent plus d’une adhésine. Posséder plusieurs types d’adhésines semble être un facteur
important dans l’établissement d’infections extraintestinales et jouer un rôle dans la virulence
de la souche (Harel, J. et al. 1991). Le gène hÏyA encode pour une hémolysine. Il a été
démontré dans un modèle d’infections urinaires de rats que les souches E. cou ne possédant
pas d’hémolysines sont moins virulentes (Hacker, J. et al. 1983). Les isolats provenant
d’humains possèdent en plus grand nombre le gène hlyA (48%) comparativement aux isolats
d’animaux (15%). Cependant, les isolats provenant d’infections urinaires animales, possèdent
en majorité ce gène. D’autres études ont aussi démontré l’importante fréquence de
l’hémolysine HIyA parmi les ExPEC isolées d’humains et d’animaux (Johuson, J. R. et al.
2000; Opal, S. M. et al. 1990). Le gène iucD appartenant à l’opéron iuc code pour un système
de captation du fer par une aérobactine. Plus d’échantillons animaux ont été positifs (69%) à
l’hybridation avec la sonde iucD que d’humains (59%). Il est intéressant de voir que la
majorité des isolats animaux provenant de tissus ont été positifs (93%) comparativement aux
isolats provenant d’échantillons urinaires animaux (7%). Peut-être que cette différence vient
du fait que ce sont deux infections différentes. Néanmoins, presque la totalité des isolats
possèdent soit le gène hlyA ou le gène iucD. Des résultats semblables ont aussi été notés par
Opal et aï. chez des isolats d’origine humaine (Opal, S. M. et al. 1990). Nous pouvons
supposer qu’en absence d’aérobactines, l’hémolysine pourrait servir de mécanisme
compensatoire.
Quatre grands groupes phylogénétiques représentant les souches E. cou, A, Bi, B2 et D, ont
été décrits par Herzer et aï. (Herzer, P. J. et al. 1990). Des analyses phylogénétiques ont
démontré que les souches virulentes extraintestinales appartiennent en majorité au groupe B2
et en faible proportion au groupe D (Boyd, E. F. and Harti, D. L. 1998; Bingen, E. et al.
199$). La plupart des souches commensales appartiennent plutôt aux groupes A et Bi. Dans
cette étude, la moitié des souches isolées d’animaux semblent appartenir aux groupes A et Bi.
126
Parmi ces isolats, la majorité est représentée par des isolats porcins. Plus de 80% des isolats
porcins ont été obtenus de porcs décédés de bactériémies primaires ou secondaires. Ceci se
démarque des résultats obtenus des isolats humains. Ceux-ci appartiennent en majorité (77%)
aux groupes B2 ou D. Cette proportion d’isolats d’origine humaine appartenant à ces groupes
a aussi été notée par d’autres études (Johnson, J. R. et aÏ. 2001; Clermont, O. et aÏ. 2000).
Selon nos résultats, il semble y avoir une différence notable au niveau du profil génotypique
de la virulence entre les isolats provenant d’animaux à celles provenant d’humains.
Les résultats phylogéniques permettre de poser l’hypothèse que les isolats testées ne sont
peut-être pas toutes des souches ExPEC proprement dites. En effet, les souches isolées
d’infections urinaires humaines possèdent plusieurs similarités avec celles isolées d’infections
urinaires animales (gènes de résistance, gènes de virulences et groupes phylogénétiques). Ce
qui n’est pas le cas entre les souches isolées d’infections urinaires et celles isolées des tissus.
Ces observations peuvent être dues au fait que les isolats provenant d’infections urinaires ont
des caractéristiques déterminées et différentes des autres ExPEC.
2.3.2 Association virulence et résistance
Il est intéressant de voir que presque aucun des gènes de virulence testés n’est associé avec un
gène de résistance. Par le test du chi-carré, seul le gène afaD8 chez l’animal, le gène sfizDE,
hlyA et iucD chez l’humain, sont associés avec la présence d’un gène de résistance (Table V
et VI, Article II). Les associations retrouvées chez l’animal entre les gènes de résistance et de
virulence sont différentes de celles retrouvées chez l’humain. Malgré cela, les associations
entre les gènes de virulence sont retrouvées à la fois chez les isolats de provenance animale et
humaine, mais à des forces différentes. Par exemple, le gène hÏyA est fortement associé avec
la présence du gène sfaDE et négativement associé avec le gène iucD. Ceci démontre que la
présence des gènes de résistance est gérée indépendamment de celle des gènes de virulence. Ànotre connaissance, cette dernière observation n’a pas été notée par d’autres auteurs. Cette
observation pourrait découler du fait que la majorité des gènes de résistance testés dans cette
étude sont retrouvés sur des éléments génétiques mobiles, ce qui les rend instables. Par contre,
les gènes de virulence qui ont été testés se retrouvent en majorité sur le chromosome
bactérien, souvent retrouvés à l’intérieur d’îlots de pathogénicité chez E. cou (Dobrindt, U. et
al. 2002).
127
Conclusions
L’étude portant sur la résistance aux agents antimicrobiens retrouvée chez différents groupes
de souches E. cou entérotoxinogènes (ETEC) 0149 : K91 isolées de porcs souffrant de
diarrhée, entre 197$ et 2000, nous permet de tirer les conclusions suivantes
1. Il y a une augmentation de la multirésistance aux agents antimicrobiens, dans le
temps, chez les souches ETEC 0149 : K9 1;
2. Pour un même phénotype de résistance aux agents antimicrobiens, différents gènes
responsables de cette résistance ont été retrouvés parmi les souches ETEC 0149
K91;
3. Les intégrons de classe 1 sont des éléments génétiques mobiles retrouvés
fréquemment parmi les souches ETEC 0149: K91. On retrouve dans la région
variable de ces intégrons des gènes de résistance aux agents antimicrobiens codant, en
majorité, pour une résistance à la tnméthoprime etlou aux aminoglycosides;
4. Certains gènes de résistance sont associés, statistiquement, entre eux, et, selon la
période d’isolement de la souche, les associations entre gènes de résistance sont
différentes.
L’étude portant sur la résistance aux agents antimicrobiens retrouvée parmi des souches
extraintestinales de E. cou, isolées de tissus infectés d’animaux et d’infections urinaires
animales et humaines, nous permet de tirer les conclusions suivantes
1. Les deux groupes partagent certaines caractéristiques quant à leur profil phénotypique
de résistance aux agents antimicrobiens;
2. La distribution de certains gènes peut être différente selon si le groupe de souches est
isolé d’animaux ou d’humains;
3. Les intégrons de classe 1 sont des éléments génétiques mobiles retrouvés
fréquemment parmi les souches E. cou isolées de tissus infectés d’animaux et
d’infections urinaires animales et humaines. On retrouve dans la région variable de ces
intégrons des gènes de résistance aux agents antimicrobiens codant, en majorité, pour
une résistance à la triméthoprime et/ou aux aminoglycosides;
128
4. Certains gènes de résistance et de virulence sont associés, statistiquement, entre eux,
et, selon l’origine d’isolement de la souche (animale ou humaine), les associations
entre gènes de résistance sont différentes.
5. Les souches isolées d’infections urinaires d’animaux présentent plusieurs
caractéristiques génétiques qui les distinguent des souches isolées de tissus infectés
d’animaux.
Dans l’avenir, il serait intéressant d’étudier dans les cas de multirésistance quelle est la
relation entre ces déterminants génétiques, s’ils appartiennent au même élément génétique de
résistance (plasmide, transposon, intégron) ou s’ils sont génétiquement indépendants. Ceci
permettrait une meilleure compréhension du mode d’acquisition des gènes de résistance.
n
SECTION IV — Bibliographie
n
n
130
Bopp, C. A., Brenner, F. W., Fields, P. I., Wells, J. G., and Strockbine, N. A. (2003).Escherichia, Shigella, and
Salmonella. In. Manuai of Clinical Microbiology. 8th. ASM press654-67 1.Wray, C., McLaren, I. M., and Carroli, P. J. (1993). Escherichia cou isolated from farm
animais in England and Wales between 1986 and 1991. The Veterinaiy record, 133:439-42.
Garabal, J. I., Gonzalez, E. A., Vazquez, F., Blanco, J., Blanco, M., and Blanco, J. E.(1996). Serogroups of Escherichia cou isolated from piglets in Spain. Veterinaiyrnicrobiology, 48: 113-23.
Wada, Y., Nakaoka, Y., Kondo, H., Nakazawa, M., and Kubo, M. (1996). Dual infectionwith attaching and effacing Escherichia cou and enterotoxigenic Escherichia cou inpost-weaning pigs. Journal ofcomparativepathology, 114: 93-9.
Nagy, B., Casey, T. A., and Moon, H. W. (1990). Phenotype and genotype ofEscherichiacou isolated from pigs with postweaning diarrhea in Hungary. Journal of clinicalrnicrobiology, 2$: 65 1-3.
Wieler, L. H., Ilieff, A., Herbst, W., Bauer, C., Vieler, E., Bauerfeind, R., et al. (2001).Prevaience of enteropathogens in suckiing and weaned piglets with diarrhoea insouthem Germany. Journal of veterinaiy medicine. B, Infectious diseases andveterinaly public health., 4$: 151-9.
Amezcua, R., Friendshïp, R. M., Dewey, C. E., Gyles, C., and Fairbrother, J. M. (2002).Presentation of postweaning Escherichia coli diarrhea in southern Ontario, prevalenceof hemolytic E. coli serogroups involved, and their antimicrobial resistance pattems.Canadian journal of veterinaly research - Revue canadienne de rechercheveterinaire., 66: 73-8.
Bischoff, K. M., White, D. G., McDermott, P. F., Zhao, S., Gaines, S., Maurer, J. J., et al.(2002). Characterization of chioramphenicol resistance in beta-hemolytic Escherichiacoli associated with dianhea in neonatal swine. Journal ofclinical microbiology, 40:389-94.
Russo, T. A., and Johnson, J. R. (2000). Proposal for a new inclusive designation forextraintestinal pathogenic isolates of Escherichia coli: ExPEC. The Journal ofinfectious diseases, 181: 1753-4.
Aubry-Damon, H., and Courvalin, P. (1999). Bacterial resistance to antimicrobial agents:selected problems in France, 1996 to 1998. Emerging infectious diseases, 5: 3 15-20.
Monroe, S., and Polk, R. (2000). Antimicrobial use and bacterial resistance. Current opinionin rnicrobiology, 3: 496-501.
Teshager, T., Herrero, I. A., Porrero, M. C., Garde, J., Moreno, M. A., and Dominguez,L. (2000). Surveillance of antimicrobial resistance in Escherichia coli strains isolatedfrom pigs at $panish siaughterhouses. International journal of antimicrobial agents,15: 137-42.
Schwarz, S., and Chaslus-Dancia, E. (2001). Use of antimicrobials in veterinary medicineand mechanisms ofresistance. Veterinary research, 32: 20 1-25.
Aarestrup, F. M., and Wegener, H. C. (1999). The effects of antibiotic usage in foodanimais on the development of antimicrobial resistance of importance for humans inCampylobacter and Escherichia coli. Microbes and infection / Institut Pasteur, 1:639-44.
Guillemot, D. (1999). Antibiotic use in humans and bacterial resistance. Current opinion inmicrobiology, 2: 494-8.
Santé Canada. Réglementation et distribution des médicaments antimicrobiens en vue deleur utilisation chez les animaux destinés à l’alimentation, http://www.hcsc.gc.ca/vetdrugs-medsvet/amr/pdf/fchapter4.pdf.
131
Alekshun, M. N., and Levy, S. B. (2000). Bacteriai drug resistance: Response to suwivaithreats. In: Bacterial stress responses. ASM Press, Washinton, D. C., 323-366.
van den Bogaard, A. E., and Stobberingh, E. E. (2000). Epidemiology of resistance toantibiotics. Links between animais and humans. Internationaljournal ofantimicrobialagents, 14: 327-35.
Ohmae, K., Yonezawa, S., and Terakado, N. (1981). R piasmid with carbadox resistancefrom Escherichia cou of porcine origin. Antimicrobial agents and chernotherapy, 19:$6-90.
Bates, J., Jordens, J. Z., and Grïffiths, D. T. (1994). Farm animais as a putative reservoirfor vancomycin-resistant enterococcai infection in man. The Journal ofantirnicrobialchemotherapy, 34: 507-14.
Coque, T. M., Tomayko, J. F., Ricke, S. C., Okhyusen, P. C., and Murray, B. E. (1996).Vancomycin-resistant enterococci from nosocomial, community, and animal sourcesin the United States. Antirnicrobial agents and chemotherapy, 40: 2605-9.
Aarestrup, F. M., Seyfarth, A. M., Emborg, H. D., Pedersen, K., Hendriksen, R. S., andBager, F. (2001). Effect of abolishment of the use of antimicrobiai agents for growthpromotion on occurrence of antimicrobiai resistance in fecai enterococci from foodanimais in Denmark. Antimicrobial agents and chemotherapy, 45: 2054-9.
Boerlin, P., Wissing, A., Aarestrup, F. M., Frey, J., and Nicolet, J. (2001). Antimicrobiaigrowth promoter ban and resistance to macroiides and vancomycin in enterococcifrom pigs. Journal of clinical microbiology, 39: 4 193-5.
Heuer, O. E., Pedersen, K., Andersen, J. S., and Madsen, M. (2002). Vancomycinresistant enterococci (VRE) in broiier flocks 5 years afier the avoparcin ban.Microbial drug resistance (Larchmont, N Y), 8: 13 3-8.
Bjorkman, J., and Andersson, D. I. (2000). The cost of antibiotic resistance from a bacteriaiperspective. Drug resistance updates . reviews and commentaries in antimicrobiatand anticancer chernotherapy, 3: 23 7-245.
Maisnier-Patin, S., Berg, O. G., Liijas, L., and Andersson, D. 1. (2002). Compensatoryadaptation to the deieterious effect of antibiotic resistance in Salmonella typhimurium.Molecular rnicrobiology, 46: 355-66.
Valenzuela, M. S., Siddiqui, K. A., and Sarkar, B. L. (1996). High expression of piasmidencoded tetracyciine resistance gene in E. cou causes a decrease in membrane-boundATPase activity. Flasmid, 36: 19-25.
Bolton, L. F., Kelley, L. C., Lee, M. D., Fedorka-Cray, P. J., and Maurer, J. J. (1999).Detection of muitidrug-resistant Salmonella enterica serotype typhimurium DT1O4based on a gene which confers cross-resistance to florfenicol and chloramphenicol.Journal ofclinical microbiology, 37: 134$-51.
Osterblad, M., Hakanen, A., Manninen, R., Leistevuo, T., Peltonen, R., Meurman, O., etal. (2000). A between-species comparison of antimicrobiai resistance in enterobacteriain fecai fora. Antimicrobial agents and chemotherapy, 44: 1479-$4.
Marshall, B., Petrowski, D., and Levy, S. B. (1990). Inter- and intraspecies spread ofEscherichia coli in a farm environment in the absence of antibiotic usage. Froceedingsofthe NationalAcademy ofSciences ofthe United States ofAmerica, 87: 6609-13.
Oppegaard, H., Steinum, T. M., and Wasteson, Y. (2001). Horizontal transfer of a multidrug resistance plasmid between coiifonn bacteria of human and bovine origin in afarm environment. Applied and environmental microbiology, 67: 3 732-4.
Hummel, R., Tschape, H., and Witte, W. (1986). Spread of plasmid-mediatednourseothricin resistance due to antibiotic use in animal husbandry. Journal of basicmicrobiology, 26: 46 1-6.
132
Hunter, J. E., Bennett, M., Hart, C. A., Shelley, J. C., and Walton, J. R. (1994).Apramycin-resistant Escherichia cou isolated from pigs and a stockrnan.EpidernioÏogy and infection, 112: 473-80.
Hunter, J. E., Hart, C. A., Shelley, J. C., Walton, J. R., and Bennett, M. (1993). Humanisolates of apramycin-resistant Escherichia cou which contain the genes for theAAC(3)TV enzyme. EpidemioÏogy and infection, 110: 253-9.
Johnson, A. P., Burns, L., Woodford, N., Ihrelfall, E. J., Naidoo, J., Cooke, E. M., et al.(1994). Gentamicin resistance in clinical isolates of Escherichia cou encoded by genesofveterinary origin. Journal ofmedicaÏ microbiology, 40: 22 1-6.
van den Bogaard, A. E. (1997). Antimicrobial resistance--relation to human and animalexposure to antibiotics. The Journal ofantirnicrobial chemotherapy, 40: 453-4.
Bertchinger, H. U., and Fairbrother, J. M. (1999). Escherichia cou Infections. In: Diseasesofswine. 8. Iowa State University Press, Ames, 43 1-468.
Fontaine, F., D’Allaire, S., Péres, S., and Fairbrother, J. M. Trends in antibiotic resistanceamong enterotoxigenic Escherichia coli serotype 0149: K91 isolated from piglets inQuébec over years. Unpublished,
Hampson, D. J. (1994). Postweaning Escherichia coli dianhoea in pigs. In: Escherichia coliin domestic animais and humans. CAB international, Guelph, 171-191.
Johnson, J. R., Steil, A. L., Delavarï, P., Murray, A. C., Kuskowsld, M., and Gaastra, W.(2001). Phylogenetic and pathotypic simiiarities between Escherichia coli isolatesfrom urinary tract infections in dogs and extraintestinal infections in humans. TheJournal of infectious diseases, 183: 897-906.
Feria, C., Machado, J., Duarte Correia, J., Goncalves, J., and Gaastra, W. (2001).Distribution of papG alleles among uropathogenic Escherichia cou isoiated fromdifferent species. FEM$ rnicrobiology letters, 202: 205-8.
Caya, F., Fairbrother, J. M., Lessard, L., and Quessy, S. (1999). Characterization of therisk to human health of pathogenic Escherichia coli isolates from chicken carcasses.Journal offood protection, 62: 74 1-6.
Johnson, J. R., and Russo, I. A. (2002). Extraintestinal pathogenic Escherichia coli: “theother bad E cou”. The Journal oflaboratoiy and clinical medicine, 139: 155-62.
Herzer, P. J., Inouye, S., Inouye, M., and Whittam, I. S. (1990). Phylogenetic distributionof branched RNA-linked multicopy single- stranded DNA among natural isolates ofEscherichia cou. Journal ofbacteriology, 172: 6175-81.
Recchia, G. D., and Hall, R. M. (1995). Gene cassettes: a new class of mobile element.Microbiology, 141: 3015-27.
Levesque, C., Piche, L., Larose, C., and Roy, P. H. (1995). PCR mapping of integronsreveals severai novei combinations of resistance genes. Antimicrobial agents andchemotherapy, 39: 185-91.
Ambler, R. P., Coulson, A. F., Frere, J. M., Ghuysen, J. M., Joris, B., Forsman, M., et al.(1991). A standard numbering scheme for the class A beta-lactamases. TheBiochemicaljournal, 276: 269-70.
Bush, K., Jacoby, G. A., and Medeiros, A. A. (1995). A functional classification scheme forbeta-lactamases and its correlation with molecular structure. Antirnicrobial agents andchernotherapy, 39: 1211-33.
Du Bois, S. K., Marriott, M. S., and Amyes, S. G. (1995). TEM- and SHV-derivedextended-spectrum beta-iactamases: reiationship between selection, structure andfunction. The Journal ofantirnicrobial chernotherapy, 35: 7-22.
Yang, Y., Rasmussen, B. A., and Shiaes, D. M. (1999). Class A beta-Ïactamases--enzymeinhibitor interactions and resistance. Fharmacology & therapeutics, 83: 141-5 1.
133
Kotra, L. P., Haddad, J., and Mobashery, S. (2000). Aminoglycosides: perspectives onmechanisms of action and resistance and strategies to counter resistance.Antimicrobiat agents and chemotherapy, 44: 3249-56.
Mingeot-Leclercq, M. P., Glupczynski, Y., and Tuikens, P. M. (1999). Aminoglycosides:activity and resistance. AntimicrobiaÏ agents and chemotherapy, 43: 727-3 7.
Shaw, K. J., Rather, P. N., Hare, R. S., and Muter, G. H. (1993). Molecular genetics ofaminoglycoside resistance genes and familial relationships of the aminoglycosidemodifying enzymes. MicrobiologicaÏ reviews, 57: 13 8-63.
Wright, G. D., Berghuis, A. M., and Mobashery, S. (1998). Aminoglycoside antibiotics.Structures, functions, and resistance. Advances in experirnentat medicine and bioÏogy,456: 27-69.
Chopra, I., and Howe, T. G. (1978). Bacterial resistance to the tetracyclines.MicrobioÏogical reviews, 42: 707-24.
Chopra, I., and Roberts, M. (2001). Tetracycline antibiotics: mode of action, applications,molecular biology, and epidemiology of bacterial resistance. MicrobioÏogy andmolecular bioÏogy reviews MMBR, 65: 232-60; second page, table of contents.
Huovinen, P., Sundstrom, L., Swedberg, G., and Skold, 0. (1995). Trimethoprim andsulfonamide resistance. Antimicrobial agents and chernotherapy, 39: 279-89.
Skold, 0. (2001). Resistance to trimethoprim and sulfonamides. Veterinaiy research, 32:261-73.
Neuman, M. (1990). Vade-Mecum des antibiotiques et agents chimiothérapiques anti-infectieux. 5. Paris.
Livermore, D. M. (1995). beta-Lactamases in laboratory and clinical resistance. Clinicalrnicrobiology reviews, 8: 557-84.
Jacoby, G. A., and Carreras, I. (1990). Activities of beta-lactam antibiotics againstEscherichia cou strains producing extended-spectrum beta-lactamases. Antimicrobialagents and chernotherapy, 34: 85 8-62.
Perilli, M., Segatore, B., De Massis, M. R., Pagani, L., Luzzaro, F., Rossolini, G. M., etal. (2002). Biochemical characterization of TEM-92 extended-spectrum betalactamase, a protein differing from TEM-52 in the signal peptide. Antimicrobialagents and chemotherapy, 46: 398 1-3.
Chang, F. Y., Siu, L. K., Fung, C. P., Huang, M. H., and Ho, M. (2001). Diversity ofSHVand TEM beta-lactamases in Ktebsiella pneumoniae: gene evolution in NorthemTaiwan and two novel beta-lactamases, SHV-25 and SHV-26. AntimicrobiaÏ agentsand chemotherapy, 45: 2407-13.
Lopez-Otsoa, F., Gallego, L., Towner, K. J., Tysail, L., Woodford, N., and Livermore, D.M. (2002). Endemic Carbapenem Resistance Associated with OXA-40Carbapenemase among Acinetobacter baumannii Isolates from a Hospital in NorthemSpain. Journal ofclinical microbiotogy, 40: 474 1-4743.
Savoie, A., Sanschagrin, F., Palzkill, T., Voyer, N., and Levesque, R. C. (2000). Structurefunction analysis of aipha-helix H4 using PSE-4 as a model enzyme representative ofclass A beta-lactamases. Frotein engineering, 13: 267-74.
Poiret, L., Gniadkowski, M., and Nordmann, P. (2002). Biochemical analysis of theceftazidime-hydrolysing extended-spectrum beta-lactamase CTX-M- 15 and of itsstructurally related beta-lactamase CTX-M-3. The Journal of antirnicrobialchemotherapy, 50: 103 1-4.
Chaibi, E. B., Sirot, D., Paul, G., and Labia, R. (1999). Inhibitor-resistant TEM betalactamases: phenotypic, genetic and biochemical characteristics. The Journal ofantimicrobiat chemotherapy, 43: 447-5 8.
134
Rasmussen, B. A., Bradford, P. A., Quinn, J. P., Wiener, J., Weinstein, R. A., and Bush,K. (1993). Genetically diverse ceftazidime-resistant isolates from a single center:biochemical and genetic characterization of TEM-lO beta- lactamases encoded bydifferent nucleotide sequences. Antimicrobial agents and chenzotherapy, 37: 1989-92.
Thomson, K. S., and Smith Moland, E. (2000). Version 2000: the new beta-lactamases ofGram-negative bacteria at the dawn of the new millennium. Microbes and infection /Institut Pasteur, 2: 1225-3 5.
Sundstrom, L., Radstrom, P., Swedberg, G., and Skold, 0. (1988). Site-specificrecombination promotes linkage between trimethoprim- and sulfonamide resistancegenes. Sequence characterization of dÏftV and sulI and a recombination active locusofTn2l. Moiecular &generai genetics MGG, 213: 191-201.
Miller, G. H., Sabatelli, F. J., Hare, R. S., Glupczynski, Y., Mackey, P., Shlaes, D., et aï.(1997). The most frequent aminoglycoside resistance mechanisms--changes with timeand geographic area: a reflection of aminoglycoside usage pattems? AminoglycosideResistance Study Groups. Clinical infectious diseases . an officiai publication of theInfectious Diseases Society ofAmerica, 24 Suppi 1: S46-62.
Sandvang, D., and Aarestrup, F. M. (2000). Characterization of aminoglycoside resistancegenes and class 1 integrons in porcine and bovine gentamicin-resistant Escherichiacou. Microbiat drug resistance (Larchmont, À’ Y), 6: 19-27.
Prescott, J. F., Baggot, J. D., and Walker, R. D. (2000). Antimicrobial therapy in veterinarymedicine. 3. Ames, Iowa.
Linde, H. J., Notka, F., Metz, M., Kochanowsld, B., Heisig, P., and Lehn, N. (2000). Invivo increase in resistance to ciprofloxacin in Escherichia cou associated withdeletion of the C-terminal part of MarR. Anti,nicrobiai agents and chemotherapy, 44:1865-8.
Cohen, S. P., Hachter, H., and Levy, S. B. (1993). Genetic and functional analysis of themultiple antibiotic resistance (mar) locus in Escherichia cou. Journal ofbacterioiogy,175: 1484-92.
Ross, J. I., Eady, E. A., Cove, J. H., and Cunliffe, W. J. (1998). 16S rRNA mutationassociated with tetracycline resistance in a gram- positive bacterium. Antimicrobialagents and chemotherapy, 42: 1702-5.
Jones, C. S., Osborne, D. J., and Stanley, J. (1992). Enterobacterial tetracycline resistancein relation to plasmid incompatibility. Moiecuiar and ceiÏular probes, 6: 313-7.
Allard, J. D., and Bertrand, K. P. (1993). Sequence ofa class E tetracycline resistance genefrom Escherichia cou and comparison of related tetracycline efflux proteins. Journalofbacteriotogy, 175: 4554-60.
White, D. G., Hudson, C., Maurer, J. J., Ayers, S., Zhao, S., Lee, M. D., et aï. (2000).Characterization of chloramphenicol and florfenicol resistance in Escherichia couassociated with bovine diarrhea. Journal ofctinicaÏ microbiology, 3$: 4593-8.
Syriopoulou, V. P., Harding, A. L., Goldmann, D. A., and Smith, A. L. (1981). In vitroantibacterial activity of fluorinated analogs of chloramphenicol and thiamphenicol.Antimicrobial agents and chernotherapy, 19: 294-7.
Murray, I. A., and Shaw, W. V. (1997). O-Acetyltransferases for chloramphenicol and othernatural products. AntimicrobiaÏ agents and chernotherapy, 41: l-6.
Bissonnette, L., Champetier, S., Buisson, J. P., and Roy, P. 11. (1991). Characterization ofthe nonenzymatic chloramphenicol resistance (cmiA) gene of the 1n4 integron ofTn1696: similarity of the product to transmembrane transport proteins. Journal ofbacterioiogy, 173: 4493-502.
Cloeckaert, A., Baucheron, S., Flaujac, G., Schwarz, S., Kehrenberg, C., Martel, J. L., etal. (2000). Plasmid-mediated florfenicol resistance encoded by the foR gene in
135
Escherichia cou isolated from cattie. Antimicrobial agents and chemotherapy, 44:2858-60.
Doublet, B., Schwarz, S., Nussbeck, E., Baucheron, S., Martel, J. L., Chaslus-Dancla, E.,et aL (2002). Molecular analysis of chromosomally florfenicol-resistant Escherichiacou isolates from France and Germany. The Journal of antimicrobial chernotherapy,49: 49-54.
Matthews, D. A., Bolin, J. I., Burridge, J. M., Filman, D. J., Voiz, K. W., and Kraut, J.(1985). Dihydrofolate reductase. The stereochemistry of inhibitor selectivity. TheJournal ofbiological chemistîy, 260: 392-9.
Flensburg, J., and Skold, 0. (1987). Massive overproduction of dihydrofolate reductase inbacteria as a response to the use of trimethoprim. European journal ofbiochernistiy /FEB$, 162: 473-6.
King, C. H., Shlaes, D. M., and Dul, M. J. (1983). Infection caused by thymidine-requiring,trimethoprim-resistant bacteria. Journal ofclinical microbiology, 1$: 79-83.
Jansson, C., and Skold, 0. (1991). Appearance of a new trimethoprim resistance gene,dÏfrL in Escherichia cou from swine. Antimicrobial agents and chemotherapy, 35:1891-9.
Jansson, C., Franklin, A., and Skold, 0. (1992). Spread of a newly found trimethoprimresistance gene, dhfrIX among porcine isolates and human pathogens. Antimicrobialagents and chemotherapy, 36: 2704-8.
Kahimeter, G. (2000). The ECO.SENS Project: a prospective, multinational, multicentreepidemiological suwey of the prevalence and antimicrobial susceptibility of urinarytract pathogens--interim report. The Journal ofantimicrobial chemotherapy, 46 Suppi1: 15-22; discussion 63-5.
Radstrom, P., and Swedberg, G. (198$). RSF1O1O and a conjugative plasmid contain suliLone of two known genes for plasmid-bome sulfonamide resistance dihydropteroatesynthase. Antimicrobial agents and chernotherapy, 32: 1684-92.
Skold, 0. (2000). Sulfonamide resistance: mechanisms and trends. Drug resistance updates.reviews and commentaries in antimicrobial and anticancer chemotherapy, 3: 155-160.
van Treeck, U., Schmidt, F., and Wiedemann, B. (1981). Molecular nature of astreptomycin and sulfonamide resistance plasmid (pBP1) prevalent in clinicalEscherichia coli strains and integration of an ampicillin resistance transposon (TnA).Antimicrobial agents and chernotherapy, 19: 37 1-80.
Radstrom, P., Swedberg, G., and Skold, 0. (1991). Genetic analyses of sulfonamideresistance and its dissemination in gram-negative bacteria illustrate new aspects of Rplasmid evolution. Antimicrobial agents and chemotherapy, 35: 1840-8.
Aarts, H. J., Boumedine, K. S., Nesme, X., and Cloeckaert, A. (2001). Molecular tools forthe characterisation of antibiotic-resistant bacteria. Veterinaiy research, 32: 363-80.
Mathew, A. G., Saxton, A. M., Upchurch, W. G., and Chattin, S. E. (1999). Multipleantibiotic resistance pattems of Escherichia cou isolates from swine farms. Appliedand environmental rnicrobiology, 65: 2770-2.
Nagy, B., and Fekete, . Z. (1999). Enterotoxigenic Escherichia coli (ETEC) in farmanimals. Veterinaiy research, 30: 259-24.
Lanz, R., Kuhnert, P., and Boerlin, P. (2003). Antimicrobial resistance and resistance genedeterminants in clinical Escherichia coli from different animal species in Switzerland.Veterina,y microbiology, 91: 73-84.
Gilmore, A. (1986). Chloramphenicol and the politics of health. CMAJ. Canadian MedicalAssociation journal - Journal de lAssociation medicale canadienne, 134: 423, 426-8,433-5.
136
Werckenthin, C., Seidi, S., Riedi, J., Kiossis, E., Wolf, G., Stolla, R., et aL (2002).Escherichia cou isolates from young calves in Bavaria: in vitro susceptibilities to 14anti-microbial agents. Journal of veterinaly medicine. B, Infectious diseases andveterinaly public heaÏth., 49: 61-5.
Sunde, M., and Sorum, 11. (1999). Characterization of integrons in Escherichia cou of thenormal intestinal fora of swine. Microbial drug resistance (Larchmont, N Y), 5: 279-87.
Mathew, A. G., Upchurch, W. G., and Chattin, S. E. (1998). Incidence of antibioticresistance in fecal Escherichia cou isolated from commercial swine farms. Journal ofanimal science, 76: 429-34.
Lee, C., Langlois, B. E., and Dawson, K. A. (1993). Detection of tetracycline resistancedeterminants in pig isolates from three herds with different histories of antimicrobialagent exposure. Applied and environmental microbiology, 59: 1467-72.
Testa, R. T., Petersen, P. J., Jacobus, N. V., Sum, P. E., Lee, V. J., and Tally, F. P.(1993). In vitro and in vivo antibacterial activities of the glycylcyclines, a new class ofsemisynthetic tetracyclines. Antimicrobial agents and chemotherapy, 37: 2270-7.
Pal, H., Lyu, S., Lee, J. H., Kim, J., Kwon, Y., Kim, J. W., et al. (1999). Survey ofextended-spectrum beta-lactamases in clinical isolates of Escherichia coli andKlebsiella pneumoniae : prevalence of TEM-52 in Korea. Journal of clinicalnicrobiology, 37: 1758-63.
Keyes, K., Hudson, C., Maurer, J. J., Ihayer, S., White, D. G., and Lee, M. D. (2000).Detection of florfenicol resistance genes in Escherichia cou isolated from sickchickens. Antimicrobial agents and chemotherapy, 44: 42 1-4.
Lee, J. C., Oh, J. Y., Cho, J. W., Park, J. C., Kim, J. M., Seol, S. Y., et al. (2001). Theprevalence of trimethoprim-resistance-conferring dihydrofolate reductase genes inurinary isolates of Escherichia cou in Korea. The Journal of antimicrobialchemotherapy, 47: 599-604.
Bass, L., Liebert, C. A., Lee, M. D., Summers, A. O., White, D. G., Ihayer, S. G., et aL(1999). Incidence and characterization of integrons, genetic elements mediatingmultiple-drug resistance, in avian Escherichia coli. Antimicrobial agents andchemotherapy, 43: 2925-9.
Zhao, S., White, D. G., Ge, B., Ayers, S., Friedman, S., English, L., et al. (2001).Identification and characterization of integron-mediated antibiotic resistance amongShiga toxin-producing Escherichia coli isolates. Applied and environmentalmicrobiology, 67: 1558-64.
Schmitz, F. J., Hafner, D., Geisel, R., Follmann, P., Kirschke, C., Verhoef, J., et al.(2001). Increased prevalence of class I integrons in Escherichia coli, Klebsiellaspecies, and Enterobacter species isolates over a 7-year period in a German universityhospital. Journal ofclinical microbiology, 39: 3724-6.
van Beikum, A., Goessens, W., van der Schee, C., Lemmens-den Toom, N., Vos, M. C.,Cornelissen, J., et aL (2001). Rapid emergence of ciprofoxacin-resistantenterobacteriaceae containing multiple gentamicin resistance-associated irttegrons ina Dutch hospital. Emerging infectious diseases, 7: 862-71.
Roy, P. H. (1997). Dissémination de la résistance aux antibiotiques : le génie génétique àl’oeuvre chez les bactéries. Médecine/$ciences, 13: 927-933.
Peters, E. D., Leverstein-van Hall, M. A., Box, A. I., Verhoef, J., and Fluit, A. C. (2001).Novel gene cassettes and integrons. Antimicrobial agents and chemotherapy, 45:2961-4.
O’Connor, A. M., Poppe, C., and McEwen, S. A. (2002). Changes in the prevalence ofresistant Escherichia cou in caffle receiving subcutaneously injectable oxytetracycline
137
in addition to in- feed chlortetracycline compared with cattie receiving only in-feedchlortetracycline. Canadian journal of veterinaly research - Revue canadienne derecherche veterinaire., 66: 145-50.
Mm, J., and Lee, H. J. (2000). Rapid discriminatory detection of genes coding for SHVbeta-lactamases by ligase chain reaction. Antirnicrobial agents and chemotherapy, 44:1860-4.
Barany, F. (1991). Genetic discase detection and DNA amplification using clonedtherrnostable ligase. Proceedings of tue National Acaderny of Sciences of the UnitedStates ofArnerica, 88: 189-93.
Rasheed, J. K., Jay, C., Metchock, B., Berkowitz, F., Weigel, L., Crellin, J., et al. (1997).Evolution of extended-spectrum beta-lactam resistance (SHV-$) in a strain ofEscherichia cou during multiple episodes of bacteremia. Antirnicrobial agents andchernotherapy, 41: 647-53.
Johnson, J. R., O’Bryan, I. T., Low, D. A., Ling, G., Delavari, P., Fasching, C., et al.(2000). Evidence of commonality between canine and human extraintestinalpathogenic Escherichia cou strains that express papG allele III. Infection andirnrnunity, 6$: 3327-36.
Johnson, J. R, Kuskowski, M. A., O’Bryan, T. T., and Maslow, J. N. (2002).Epidemiological correlates of virulence genotype and phylogenetic background amongEscherichia coli blood isolates from adults with diverse-source bacteremia. TheJournal of infectious diseases, 185: 1439-47.
Chanal-Claris, C., Sirot, D., Bret, L., Chatron, P., Labia, R., and Sirot, J. (1997). Novelextended-spectrum TEM-type beta-lactamase from an Escherichia cou isolateresistant to ceftazidime and susceptible to cephalothin. Antimicrobial agents andche,notherapy, 41: 7 15-6.
Oluoch, A. O., Mm, C. H., Weisiger, R. M., Koo, H. Y., Siegel, A. M., Campbell, K. L., etal. (2001). Nonenteric Escherichia cou isolates from dogs: 674 cases (1990-1998).Journal ofthe American Veterinary Medical Association, 21$: 38 1-4.
Feria, C., Ferreira, E., Correia, J. D., Goncalves, J., and Canica, M. (2002). Pattems andmechanisms of resistance to beta-lactams and beta- lactamase inhibitors inuropathogenic Escherichia cou isolated from dogs in Portugal. The Journal ofantimicrobial chernotherapy, 49: 77-85.
Bradford, P. A., Petersen, P. J., Fingerman, I. M., and White, D. G. (1999).Characterization of expanded-spectrum cephalosporin resistance in E. cou isolatesassociated with bovine caif diarrhoeal disease. The Journal of antimicrobialchemotherapy, 44: 607-10.
van den Bogaard, A. E., London, N., Driessen, C., and Stobberingh, E. E. (2001).Antibiotic resistance of faecal Escherichia cou in poultry, poultry farmers and poultryslaughterers. The Journal ofantimicrobial chernotherapy, 47: 763-71.
Smyth, C. J., Marron, M., and Smith, S. G. J. (1994). Fimbriae of Escherichia cou. In:Escherichia coli in domestic animals and humans. CAB international, Guelph, 399-435.
Hard, J., Daigle, F., Maiti, S., Desautels, C., Labigne, A., and Fairbrother, J. M. (1991).Occurrence ofpap-, sfa-, and afa-related sequences among F165-positive Escherichiacou from diseased animals. FEMS ,nicrobiouogy letters, 66: 177-$2.
Hacker, J., Hughes, C., Hof, H., and Goebel, W. (1983). Cloned hemolysin genes fromEscherichia cou that cause urinary tract infection determine different levels of toxicityin mice. Infection and irnlnunity, 42: 57-63.
SECTION V - Annexes
Période(Années)
ooo(J‘t,O)O)
xxii
Annexe 1: Évolution de la multirésistance chez les souches Escherichia cou
entérotoxinogènes
Résultats
À la suite de l’analyse des antibiogrammes effectués sur les isolats ETEC, nous pouvons
remarquer que les souches isolées entre 1978 et 1984 sont résistantes en majorité à moins de
trois agents antirnicrobiens. Nous pouvons faire les mêmes observations pour les souches
isolées entre 1985 et 1989. Par contre, dès 1990, le nombre de souches résistantes à plus de
trois agents devient plus élevé. Et entre 1995 et 2000, les isolats ETEC sont en majorité
résistants à plus de trois agents antimicrobiens allant jusqu’à huit résistances. Il y a, dans le
temps, un phénomène d’ascension de la multirésistance chez les souches ETEC.
wC.,
e(nw
n
Nombre de résistancesphénotypiques
Figure 2. Évolution ascendante de la multirésistance chez les souches ETEC isolées de porcs
atteints d’une diarrhée postsevrage entre les années 197$ et 2000
xxiii
Annexe 2: Analyse statistique du chi-carré entre différents gènes de résistance et
l’intégron de classe 1 retrouvés parmi les souches ETEC isolées entre 1978 et 1989
Résultats
Voir article L
Tableau III. Association des gènes de résistance et de l’intégron de classe 1 parmi les
souches ETEC isolées entre 197$ et 1989
Gènes de résistance aux agents antirricrobiens et intégron de dasse la
7Etf Si-1V izAl 42 ‘(3-iV chfl chI5 &(r13 cUtIS stil suS tatA tatB tatC cati
Si-iv -
L4I + -
-
ca(J)-IV + - ++ -
ci*l - - - - -
cHat - ++ - - - -
cttrl3 - - - - - - -
cUrI5 - - - - - +++ - -
stil + - - - - - - - -
scSi - - - - - - - + - -
tatA + + + - - - - - - - -
ttis + + - - - - - - - - - (+.H-)5
lelC + + + - - - - - - - - +++ (+-l-+)
cati - - - - - - - - - + - -
51 - - - - - - - - - - - -
Seuls les gènes de résistance aux agents antimicrobiens qui montraient une association avec
un autre gène de résistance à niveau F < 0.05 sont inscrits. Les niveaux significatifs
d’associations (tels que donnés par le test du chi-carré exact): -‘ P> 0.05; +, 0.05 = P = 0.01;
++, 0.01 F 0.001; +++, 0.001 = F.h Les parenthèses déterminent les associations négatives.
xxiv
Annexe 3: Analyse statistique du chi-carré entre différents gènes de résistance et
l’iutégron de classe 1 retrouvés parmi les souches ETEC isolées entre 1990 et 2000
Résultats
Voir article I,
Tableau W. Association des gènes de résistance et de l’intégron de classe 1 parmi les
souches ETEC isolées entre 1990 et 2000
Gènes de résistanœ aux agents antimicroens et ntégron de classe 1a
s’m slw Ml aphA2 aac(3-lV chfi dh,f5 dhfrl3 dlfrf5 su!! s’il!! tetA !eB teC cati
shv ++
apM! ++ -
aphA2 - - -
aac(39-IV ++ - + -
dhfrl - - - - -
dhfril + ++ - - + -
dhfrl3 - - - - - - -
dhfrf5 - - - - - +++ - -
s’il - - - - - - - - -
suffi - - - - - - ++ - -
tfflA - + - - - - - -- -
tets - - - - - - - -- ++ - (+H)
teIC - + - - - - - - - (++) - +++
oeil + - - - - - + - - - -
cI1 - - - - - - -
- - - - IiregrQ1
Seuls les gènes de résistance aux agents antimicrobiens qui montraient une association avec
un autre gène de résistance à niveau P < 0.05 sont inscrits. Les niveaux significatifs
d’associations (tels que donnés par le test du du-carré exact): -, P> 0.05; +, 0.05 = P = 0.01;
++, 0.01 = P 0.001; +++, 0.00 1 P.b Les parenthèses déterminent les associations négatives.
xxv
n Annexe 4: Distribution du phénotype de résistance aux agents antimicrobiens selon
l’âge des porcs
Résultats
Les souches isolées de porcs âgés entre 1 et 10 jours possèdent en majorité moins de trois
résistances. Les souches isolées de porcs âgés entre 11 et 20 jours possèdent deux ou quatre
résistances, et ce, en majorité aux agents antimicrobiens. Les autres isolats provenant de porcs
âgés entre 21 et 70 jours qui ont été isolées possèdent un nombre de résistances plus varié
entre elles, mais pouvant atteindre un nombre plus élevé que six. L’âge des porcs semble être
un facteur qui interfère dans la multirésistance retrouvée chez les souches E. cou
entérotoxinogènes.
U)
zoU)w
Âge des porcs(jours)
Figure 3. Pourcentage de souches ETEC selon le nombre de résistances aux agents
antirnicrobiens testés en fonction de l’âge des porcs d’où proviennent les souches
23
4Nombre de résistances
phénotypiques8 1-10
xxvi
Annexe 5 : Distribution de l’âge des porcs selon l’année d’isolement de la souche
Résultats
Entre 1978 et 1984, plus de 70% des souches ETEC étaient isolées de porcs âgés entre 1 et 20
jours. Entre 1985 et 1989, les souches ETEC de notre collection étaient isolées de porcs avec
un ratio égal entre les quatre différents groupes d’âges. La même observation est notée pour
les souches isolées entre 1990 et 1994. Cependant, entre 1995 et 2000, les isolats proviennent
surtout (84%) de porcs âgés entre 21 et 70 jours.
U)w
zoU)wV
11-20
Âge des porcs(jours)
1995-
1990- 2000
1994
Période d’isolement dessouches (années)
31-70 1978-1984
Figure 4. Pourcentage de souches ETEC selon de l’âge des porcs d’où proviennent les
souches en fonction de la période d’isolement de celles-ci
DD
ï)
IDENTIFICATION DE L’ETUDIANT
Nom de Féttidiant- : Code permanent
Christnè MaynardSigle du programme Titre citi programme Optinn
M .Sc. Sciences vétrinaire Micwhiologi
DESCRIPTION DE L’ARTICLE
Déclaration ,.-
. -—
‘. - 3’
À titre de coauteur de l’article identifié ci-dessus, j’autorise le microfilmage du mémoire etje suis d’accord que
Christine Maynard inclut cet article dans son mémoire de maîtrise qui a pour titre « Distribution des gènes de
résistance aux agents antimicrobiens parmi des souches Escherichia cou entérotoxinogènes et extraintestinales ».
____________________
ty t63/Ô3
_____________
2o )o57v3
o(1 O
Coauteur —
Signature-
DaterL
Cçauteur Signaturj:-.
Coauteù.. Signturé.
.. Signature -:‘:.j,-. -
I isI2d3ovY
t ACCORD ET PERMISSION DES COAUTEURS D’UN ARTICLE’ I
Auteurs L-
Christine Mayiard, John M. Fairbrother, Sadjia Bekal-Si Ah, François Sanschagrin, Roger C. Levesque, Roland Brousseau,
Luke Masson, Serge Larivière et Josée Haret
Titre “‘
Characterization and detection ofantimicrobial resistance genes in enterotoxigenic Escherichia cou 0149: K91 isolated from
pigs over a 23-year period
Revue ---.-‘ ‘-. ‘ ptedepu:.
Soumis à Antimicrobial Agents and Chemotherapy Indéterminée
DEcLARATI0N DES COAUTEURS
oautif -
Coaittir - . . -.
L4J]Coiitcur‘-.-X-\ct - Ç r1L I,
Coatitetir
Dâté ,
7-:L3
‘Annexe II du Guide de présentation et d’évaluation des mémoires de maîIrisc’ et des thèses de doctorat, mars 2001FMV-SAÉ 2001 06
0-1/25/03 11:05 FAX 511 196 6213 CNRC IRB MOL.BI0 li 002
fl ACCORD ET PERÏ\’IISSION DES COAUTEURS D’UN ARTICLE’ Ï
IDENTIFICATION DE L’ETUDIA?T
Nom de I’ftudisnt .: : .S Cnde permanent
Christinc Mavnard
Sigle du programme Titre du piogiammo Option
M. Sc. Sciences vétérinaires Microbiologie
DKc;iuvnox DE L’ARTICLE
DECLARATION DES COAUTEURS
Déciai-ation
A [itrc & uoauteur de l’article idcutflé ci-dessus. j’auturisc L inierofi]magc du mémoire etje suis d’accord que
Christine Maynard inclut cet article dans son mémoire de maîtrise qui a pour titre Distribution des gènes dc
résistancc aux agents antimicrohiens par-mi des souches £sche,ichio cou entrotoxmognes [ cxtrairitctinales >.
Çonuteur•: ::::: •: ..,,,:. ‘::.Sign
i;”’ Si
:;,,, :1 ‘:.s ::.
444q11,bo 2’2/OL(Or
CvduteufD9te
Coauteur-i .
‘:::..:..:. Si.atrne • :• -
- .. Date:,
Coauteur Signature Date
Co.iuticr SLgnature Date
Coauteur Sgnnture Date
AuteursChristinc Maynrd, John M. Fairbrottier, Sadjia Bekal-Si Ah, François Sanschagtin, Rcgcr C. Lcvcsquc. Roland Brousseau,
Luke Masson. Scrge Lariièrc et Josée Hare—
‘Titre I
Characferi7.aticrn and detecLion of anunilcrobial resistance gcncs in enterotoxienic çchr-ichie co!i 0149. K91 isulated froin
pies over n 23-vear period
ResueDate de puhiccatron
Soumis r Anrnicmohia1 Agents and Chemothcnipy —Indéterminée
%,r:rcmc 1f d C;- dr pr &ntu:ion ct ri c ntrrt,: Jç I,:é,,,orr(s ,,:i’t.se et der iri..x d dr irru ,r.,:,, 2’)’FM\’-SAE 2C O
03/18/03 11:17 FU 514 496 6213 CNRC IRB MOL.BI0
t ÀoeORD ET PERMISSION DES COAUTEURS D’UN ARTTCLE11
IDENTifICATiON DE L’ETUDIANT
.;«L 71j li ijw,.;J, ,ç fl:.Cti
M.Sc. Sciences vétérinaires . Microbiologie
DEscifirnoN DE L’ARTICLE
Soumis à Antimicrobial Agents and Chemothcrupy Indéterminée --
DECLÀa4TI0N DES COAUTEURS
À titrc de coauteur de l’article identifié ci-dessus, j’autorise le rtdcrofllmage du mémoire etje suis d’accord que
Christine Mayriard inclut cet article dans son mémoire de maitrise qui a pour titre «Distribution dcs gènes de
résistance aux agents antimicrobiens parmi des souchos Esch crichia ccli cntérotoxinogènes et extaintestinales ».
Ro2rctcYŒr\aeçàiliL 41 ii ‘M N44q ijj eflj’ig
nÇ.S&t1fl (P—?-
t r
,
t
ti2(uChnstrne Maynard — -
—
Christine Maynard, John M. Fairbrother, Sadjia Bekal-Si Mi, Ynnis Snschagriu, Roger C. Levesquc, Roland Brousseau,
Luke Massou, Serge Lerivièrc ct Josée Hard
L43, ‘q’ U cWt’ R j ‘j ‘“_ tta.n j,jj k’ iL g, iti4t,i L,iLbkq.i&&U liQl’d rtCharacterization and delechon cf antimierobisi resistance genes in cntcrotœdgenic Ercherichia ccli 0149; K91 isolated freinpigs avec a 23-year perlai
i;:.
2i
‘Annexe!! du Guide tkprdsentation et d Wa!uartôn des nzè,’;oires de ,rnUUrjs’ et de, thèses de doctôrar. mars 2001flvW-SAS 2001 06
j ACCORD ET PERMISSION DES COAUTEURS D’UN ARTICLEq
IDENTIFICATION DE L’ETUDIANT
Nom de l’étudiant Codepermanent
Christine MaynardSigle du programme Titre du programme . Option -
-
M. Sc. Sciences vétérinaires Microbiologie
DESCRIPTION DE L’ARTICLE
Auteurs —
—
Christine Maynard, Sadjia Bekal-Si Ah, François Sanschagrin, Roger C. Levesque, Roland Brousseau, Luke Masson, Serge
Larivière et Josée Harel
TitreGenetic analysis ofantimicrobial resistance among extraintestinal Escherichia cou isolates of animal and human origin
RevueDate de publication
Soumis à The Journal ofAntimicrobial Chemotherapy Indéterminée
DECLARATION DES COAUTEURS
Déclaratih”: ,..
À titre de coauteur de l’article identifié ci-dessus,j’autorise le microfilmage du mémoire etje suis d’accord que
Christine Maynard inclut cet article dans son mémoire de maîtrise qui a pour titre «Distribution des gènes de
résistance aux agents antimicrobiens parmi des souches Escherichia cou entérotoxinogènes et extraintestinales ».
Co te r:. .
. .i%•/P
Coaute r .
- . Date
L 1ASfACoauteur
-Date
Coauteur - . i- --. -,.- Signa ‘- Date. . t
Coauteur .Sig z- . - . :::
z
Coauteur . --: ..
Signatûrç
Coauteur Signature Date
Coauteur.
.“ -i-: - Signature
Envoyé a lâ FES le
1An,iexe II du Guide de présentation et dévaluation des mémoires de maîtrise et des thèses de doctorat, mars 2001FMV-SAE 2001 06
03/18/03 11:15 FAX 514 496 6213 CNRC IRB f0L.BI0 L1002
I ACCORD ET PERMISSION DES COAUTEURS D’UN A1{TICLÉ’ 1
IDENTifICATION 0E L’ETUDL4NT
Christine MaTiardI ,
M.Sc. I Science.s vétérinaires - Microbiologie
DESCRIVI’ION DE L’ARTICLE
__: t
Soumis à The Iuurnal cf Antitnicrobial Cbemotherapy I Indéterminée
DECLARÀTION 0FS COAUTEURS
À ULre de coauteur dc 1’aidc1e identifié ci-dessus, j’autorisc le microfilmage du mémoire etje suis d’accord que
Chritine Maynard inclut cet article dnns son mémoire de iuaitrise qui a pour titrc « Distiiliution des gènes de
réistançc aux agents antimiorobiens parmi des s chcs Escherichia cati entérctoxinogàncs et exlraintestinales ».
f?rr C Lvrqie1? -
:ci--r- &3-3
•‘- ,:
..
Christine Maynard, Sadjia Bekal-Si Ah, Frrnçois Satscbain Roger C. Levesque, Roland Erousseau3 Luke Mason, SergeLarivière et Josée Harel
ii rV ijt IiOenetic analysis of antin2icrobial rcistance amng exfraintestinai Eshrichfa ccli iso1ate of animal and buman engin
—01
1Am Jidu cujdt dpr n(aic,1 cf d’éa1uatiori ck: minir dt maîtHs dex thd.es de di.eWrjt. mars 200!FMV-SA 2001 06
0-1/25/03 11:06 fÀL 514 496 6213 CNRC IRB MOL.BIO 0fl3
LAccORD ET PERMISSION DES COEUTEURS D’UN ARTICLW 1
IDENTIFICATION DE L’ETUDIANT
Nôzu & rétt±i,:’:;’,::.,.... .‘‘‘. ..:‘..:.. .....
•‘.‘..,.Code permanent
Christinc Maynard —
Sigle du programme fitre dii piogrammt Option
M.Sc. Sciences vétérïnaires -Microbiologie
I»SCRIPTI0N DE L’ARTICLE
Auteurs
Christine Maynurd, Sadjia Bekal-Si Ail, François Saschagrin, Rocr C. Levesque, Rohrnd Brousseau, Luke Masson, Serc
Larivière et Josée Hard
fitre
I
Genetic analysis ofantirnictobiat rcsistancc among ex:aintestiria1 Escheric*io ccii isotates ofaiinial and hutnan origin
Re ue -
Date de pub]ication
Soumis à The Journal ofAntirnicrobial Chernothcrapy-
Indéterminée
DECLARATION DES COAUTEURS
Déclai-iriti.%.’ :.:.: •.,.,:...,., f.......... f;,..:. ‘•‘“.:
À tit.rc de coauteur de l’article identifié ci-dessus,j’autorisc Ir iriicrofilmage du mémoire etje suis d’aeord que
Christine Maynard inclut cet article dans son n:émoire dc maittise qui a pour titre « Distribution des gènes de
résistance aux agents antimirrobiens parmi des souches £schcrichict coi! entérotoxinogénes et extiaintestinales ».
Cùauteuï SignatuDate
Caut?uT Sigwitu
Date
SERGE LARIVIERE//Ç
:‘ •‘:.
‘;:. Signatu h.. .Da .i..::;;:: —
Caùtetr’*-:’’Zf’.. ‘.‘
Sigiatïe ::‘.J. ‘:..: ‘*‘ H;.»:
.Cotiteurf ... ‘. f: H’. Signature f ; ..“‘,‘,: . ,:
Coautcu.i SignatureX)qtc
ÇM9uteur -
Signature Date
Ciiteùr:1;
-. , ‘‘ .,...:..
‘,
Signture f:•. :,, . ..:
:. ,.: . Dati
Lnvv a1a FES le 2o3 ô
‘A,’,vc J! dt Quick pren:o’I cr S d,z.ç trwirt d ,,icîtr:sc e( &x ihit- ‘1g- d,ctor.jt ,,:aus 2û01fMV-S!±